Pega Platform - Architecture and Technical Brief - 0

Trademarks
For Pegasystems Inc. trademarks and registered trademarks, all rights reserved. All other trademarks or service
marks are property of their respective holders.
For information about the third-party software that is delivered with the product, refer to the third-party license
file on the installation media that is specific to the release.
Notices
This publication describes and/or represents products and services of Pegasystems Inc. It may contain trade
secrets and proprietary information that are protected by various federal, state, and international laws, and
distributed under licenses restricting their use, copying, modification, distribution, or transmittal in any form
without prior written authorization of Pegasystems Inc.
This publication is current as of the date of publication only. Changes to the publication may be made from time
to time at the discretion of Pegasystems Inc. This publication remains the property of Pegasystems Inc. and
must be returned to it upon request. This publication does not imply any commitment to offer or deliver the
products or services described herein.
This publication may include references to Pegasystems Inc. product features that have not yet been licensed
by the organization Address. Address questions about whether a capability is included in the installation, please
consult Pegasystems Inc.
Although Pegasystems Inc. strives for accuracy in its publications, any publication may contain accuracies or
typographical errors, as well as technical inaccuracies. Pegasystems Inc. shall not be liable for technical or
editorial errors or omissions contained herein. Pegasystems Inc. may make improvements and/or changes to
the publication at any time without notice.
Any references in this publication to non-Pegasystems websites are provided for convenience only and do not
serve as an endorsement of these websites. The materials at these websites are not part of the material for
Pegasystems products and use of those websites is at your own risk.
Information concerning non-Pegasystems products was obtained from the suppliers of those products, their
publications, or other publicly available sources. Address questions about non-Pegasystems products to the
suppliers of those products.
This publication may contain examples used in daily business operations that include the names of people,
companies, products, and other third-party publications. Such examples are fictitious and any similarity to the
names or other data used by an actual business enterprise or individual is coincidental.
This publication provides a technical overview of Pega’s capabilities and architecture. The exact capabilities
available to vary based on contract and license terms. Please contact the applicable Pega account executive for
more details.
This document is the property of:

Pegasystems Inc.
One Rogers Street, Cambridge, MA 02142-1209, USA
Phone: 617-374-9600 Fax: 617-374-9620
www.pega.com
Document: The Pega Platform: Architecture and Technical Brief
Publication date: 2020
© Copyright 2020 Pegasystems Inc. Confidential All rights reserved. All trademarks are the property of their respective owners. 1
Table of Contents
Trademarks ...................................................................................................................... 1
Notices .............................................................................................................................. 1
Abstract ............................................................................................................................... 4
Section 1: An Introduction to Pega Platform ................................................................... 5
Section 2: Low-code application development ............................................................... 9
2.1 An example of a model-based strategic business application .......................... 12
2.2 Defining an application and grouping models .................................................... 14
2.3 Situation-based modeling ..................................................................................... 16
2.4 Decision management and models...................................................................... 19
2.5 Declarative computation models ......................................................................... 20
2.6 Design validation .................................................................................................... 21
Section 3: Pega Platform unified, open architecture and dynamic run time ............. 22
3.1 Overview of the Pega Platform unified architecture and dynamic run time ... 22
3.2 Dynamic model selection ...................................................................................... 23
3.3 Dynamic working memory, caching, and persistence ........................................ 26
Section 4: Pega Platform dynamic user experience ..................................................... 28
4.1 Dynamic user experience ...................................................................................... 29
4.2 Mobile apps ............................................................................................................ 33
4.3 Mobile offline support ........................................................................................... 34
4.4 Intelligent Virtual Assistant .................................................................................... 35
4.5 Email bot ................................................................................................................. 37
4.6 Integrated social collaboration ............................................................................. 37
4.7 The Digital Experience (DX) API ............................................................................. 39
4.8 Pega Process Extender for Salesforce Lightning ................................................. 39
4.9 Pega Web Mashup ................................................................................................. 40
Section 5: Pega integration capabilities ......................................................................... 41
5.1 Pega® Data Integration .......................................................................................... 41
5.2 Pega Live Data ........................................................................................................ 43
5.3 Pega Robotic Process Automation ....................................................................... 44
Section 6: Pega Platform and artificial intelligence ....................................................... 46
Section 7: Cloud at Pega .................................................................................................. 49
7.1 Cloud Choice ........................................................................................................... 49
7.2 Pega Cloud Services ............................................................................................... 50
7.3 Client-managed cloud ............................................................................................ 51
Section 8: Pega Platform enterprise-class applications ............................................... 52
8.1 Enterprise-class security ........................................................................................ 52
8.2 Enterprise-class deployment and manageability in the cloud or on-premises53
8.3 Enterprise-class reliability and scalability ............................................................ 54
8.4 DevOps and automated testing ............................................................................ 55
Section 9: Pega Platform — digital transformation ...................................................... 57
9.1 The broader capabilities of Pega’s applications .................................................. 57
Conclusion ........................................................................................................................ 59
Abstract
Pega Platform™ is a low-code case management and application development platform intended for
enterprises seeking to build, deploy, and evolve strategic business applications. Strategic business
applications are those that embody the key differentiators and innovations for the activities of an
enterprise1. Whether deployed on-premises or in the cloud, these are applications that need to evolve and
continually improve as the enterprise, its customers, technology, and its competitive environment change.
Pega Platform enables the development and deployment of strategic intelligent automation, case
management, and decision management applications through a single, unified model-based platform.
Pega Platform employs an entirely model-based – or low-code – application development approach and
dynamic run-time architecture in order to provide both the agility and productivity of model-based
development as well as optimized run-time performance. Through this low-code development model,
organizations are to improve business and IT collaboration throughout the development process,
delivering high-quality applications, fast.
Pega Platform supports responsive, dynamic, and situation-based user experiences across all devices and
channels for end-user applications. Gartner calls this capability “multiexperiences.” Applications can be
specialized by multiple dimensions to meet the varied needs of multiple geographies, customer segments,
product lines, and channels. This capability enables application reuse and simplifies the overall
management and maintenance of the application. Unique reusable Pega technology scales from initial
projects to enterprise-wide programs and reduces the total cost of ownership of applications that meet
the needs of diverse business units, products, and customers. Enterprise-class Quality-of-Service (QoS)
features ensure secure, scalable, reliable, and manageable production performance for both Pega Cloud®
and deployment on “client-managed” clouds, including on-premises, private clouds.
1 “What is Strategy,” Michael E. Porter, Harvard Business Review, reprint 96608.
Section 1: An Introduction to Pega Platform
Pega Platform provides enterprises a robust low-code development platform, focused on case- and
process-centric applications. Pega Platform helps automate, integrate, and improve strategic front- and
back-office business operations, unifying customer engagement with business objectives. It is a
consolidated platform with industry-leading capabilities in business process management (BPM),
intelligent automation (IA), robotic process automation (RPA), case management2, analytics-driven decision
management (DM) and machine learning, mobility and user experience (UX), and social collaboration.
Figure 1: Pega Platform
“Unified” means that Pega Platform is not a conglomeration of separate systems with differing
development and deployment approaches. Pega Platform has a singular model-based approach to
application development and deployment in which all models that define the application are uniformly
built, managed, deployed, and monitored. Whether an application capability is considered to be part of
Intelligent Automation (also known as Digital Process Automation or DPA), BPM, DCM, DM, UX, or
integration, the models can be mixed and customized to meet the needs of the enterprise.
This also ensures that applications behave the same regardless of underlying architecture. Whether Pega
Platform is deployed in client-managed clouds or on-premises, or it is leveraged within the fully managed
Pega Cloud® Services, the application experience is the same.
2Don Schuerman, Ken Schwarz, and Bruce Williams, Dynamic Case Management for Dummies
http://www.pega.com/resources/dynamic-case-management-for-dummies, John Wiley & Sons, 2014
“Omnichannel” or “multiexperience” means that the same experience, capabilities, analytics, and content
can be accessed and used across all interaction channels and devices. A user can start an interaction in
one channel (social media, email, text, chat, etc.) and switch to another midway through the process, and
the entire interaction will be reflected in the new channel. For example, a customer can start a product
selection on a mobile device, and then complete it on the desktop by using web self-service or on the phone
with a customer service representative (CSR). No matter how many channels are used, Pega Platform
maintains the context of the interactions for all participants.
“Model-driven” or “low-code” application development provides outstanding productivity and collaboration

between the business and technology teams. The visual definition of an application component at a level
higher, less technical, and closer to business semantics makes it easier for the business to get involved in
the development of the application. It enables IT to effectively communicate with the business and
development to get a production-ready application fast. Unlike code, the visual models are accessible to
both technology and business staff, providing a common platform for communication. Like code, models
can be shared, versioned, and used in an agile development process.
The Pega Platform model-based approach enables applications to be easily built to support large numbers
of specializations according to factors such as customer status, product type, and geography, while also
providing significant reuse and deployment flexibility. These applications can also evolve in an agile and
incremental manner.
Figure 2: Situational Layer Cake™
Figure 2 shows the Pega Platform multdimensional architecture – named Situational Layer Cake™ – in
action, with sets of models organized in versioned layers. The Situational Layer Cake enables enterprises
to differentiate, specialize, and reuse their business applications based on organizational needs. Common
models are located in the lower layers while new functionality or specializations reside in the higher layers.
This structure allows for simple application reuse while also enabling each business area to build the
specializations they need for strategic differentiation for their market, products, and customer segments,
for example.
Every time the application interacts with a user or advances a case or process, Pega Platform dynamically
selects the policy and procedure that are most relevant to the situation at hand. This is especially useful
for the rapid creation and deployment of pilots or the use of A-B Testing to vet new capabilities before
releasing to the masses. A specific set of pilot users can be designated to experience the pilot, while other
users continue with their existing version.
To best leverage collaboration between the business and technology teams, Pega Platform structures each
application by the “three pillars” that underpin the Pega Express™ design thinking methodology.
1. Microjourneys – the smaller segments of a broader customer journey that lead to meaningful
outcomes – as a case, which models work that is to be done in a way that is intuitive to the business.
A case is a unit of work intended to achieve a business outcome, such as resolving a customer
service request, fulfilling an order, or completing an order-to-cash process. The case captures the
life cycle of work from inception to resolution. A case can include both structured and ad hoc
processes and simplifies the communication and development, streamlining the collaboration
between the business and IT. Additionally, a case makes it easier for the business to communicate
with their customers as each step within the process is outlined, setting the correct expectations
and improving customer satisfaction.
2. Personas and channels. These are users and stakeholders associated with an application, and the
ways in which they will access, such as web portals, mobile apps, or coming in via an API.
3. Data and interfaces. This is the data the case needs to complete the work, and interface logic for
retrieving that data from external systems and mapping into a standardized business format for
the case to use.
Figure 3: The three pillars
Figure 3 shows the case model for a mortgage application at the top. The case consists of a set of stages
across the top that track the case through its life cycle. Beneath each stage are steps that represent the
business activities to be performed within that stage. After a case is defined at the level of stages and steps,
the more detailed application models can be built in the context of the high-level case model. Because the
high-level model is universally understood, its use as the design center facilitates effective communication
and business/IT collaboration. Next, you can see the personas and channels used in the case, and below
that, the data and interfaces.
The key to high productivity in Pega Platform is that low-code models can be used to build the entire
application while still having efficient executable code automatically generated and executed directly from
the models. This approach removes any need to develop at the programming language level. Changes to
models can be incrementally deployed without a need to “rebuild” the application for deployment.
Deployment of Pega Platform applications can happen in Pega Cloud, in a client-managed cloud, or on-
premises, including in private clouds. Clients can move applications deployed in one environment to
another, for example, migrating an on-premises application to Pega Cloud. Quality of service is provided
with scalable and high-availability architectures and because the execution code is generated, best-practice
security is built into the application.
Pega Cloud provides a virtual private cloud for the Pega Platform applications. It leverages the cost, reach,
and availability benefits of a public cloud while providing a secure infrastructure for each enterprise
customer to assure protection, compliance, and performance. Pega Cloud is a fully managed service. While
Pega Cloud instances run on Amazon Web Services (AWS) infrastructure, Pega provides for management,
upgrades, and support of both the application and infrastructure layers.
Pega Robotic Process Automation™, which is included with Pega Platform, enables organizations to
automate specific work tasks through desktop robots or “bots.” The bots can be deployed as attended
desktop robots that assist users as they complete their work (RPA “attended”) or as an unattended robot
that works independently in the back office.
Section 2: Low-code application development
Using low-code models for application definition facilitates collaboration between the business and
technology teams through a common visual language and visual representations.
Each aspect of a Pega Platform application is entirely defined through visual models. Models range from
fine-grained (a single data item) to fully defined case flows. Models define every process, rule, UI form or
element, data, back-end integration, analytics, etc. – every capability in Pega Platform is available as a
configurable model. Each of these models can be reused in additional applications and contexts, but more
importantly, any model can be specialized and extended for other uses. Models are easily customized to
support large numbers of variations according to factors such as customer status, product type, and
geography.
Four distinct workspaces – App Studio, Dev Studio, Admin Studio, and Prediction Studio – allow for the
development and management of applications in a way that enables team members to focus on the
tasks that align with their expertise.
• App Studio is the primary low-code workspace, allowing both business and technical developers to
define application cases, UX, and data models without code, and enabling communication
between the business and the developer. App Studio guides users though the three pillars of an
application: microjourneys, personas and channels, and data and interfaces.
• Dev Studio allows for more advanced development functions, such as configuring integration
mapping or access controls. Rulesets and other components can be built here and made available
for use in App Studio.
• Admin Studio provides administrators the ability to manage applications, data, users, etc.
• Prediction Studio allows data scientists and analysts to define and tune predictive and machine
learning models.
When needed, users can seamlessly toggle from one studio to another via a drop-down menu.
App Studio
Dev Studio
Admin Studio
Prediction Studio
Pega Platform supports the automatic generation of documentation from the models for review,
collaboration, and compliance. The documentation includes inventories of models as well as visual
elements such as flows and entity-relationship diagrams for data models. Objectives and specifications
entered during the models’ creation and updates are included, providing a versioned history of the
supporting information for the models and applications. The documentation can serve as an administrative
guide for training purposes as well.
2.1 An example of a model-based strategic business application
To make the discussion more concrete, this section introduces an example to be used through the rest of
this paper. It represents a common use case for Pega Platform customers in the financial industry. The
capabilities of Pega Platform demonstrated in the examples all have analogs in other industry verticals
such as insurance, healthcare, telecommunications, energy, government, life sciences, and manufacturing.
Consider the needs in a retail bank for selling and providing a new product to a customer, who might be
an existing or a new customer. This process is called “customer onboarding” and is critical to both revenue
generation and customer satisfaction. The example Pega Platform application provides the front-office
application for use by both the customer and by bank employees. The focus here will be on a mortgage
product, that is, the case when the customer desires a new mortgage, and the bank’s objective is to
ultimately issue the mortgage.
ONBOARDING FOR RETAIL BANKING APPLICATION
Stages of Mortgage Product Case

1. Assess needs
Mobile
- Price and down payment?
- Also selling a house?
2. Customer verification
- Conduct credit and other checks
Web
3. Product Selection Bank Call
Center Or
- Fixed or adjustable mortgage?
Branch
- Term?
Telephone - Present prioritized recommendation
4. Collect additional customer details
- Employment and income?
Customer
- Assets and liabilities? Service
Retail Banking Email - Collect documents Rep.
Customer
5. Configure application and process (CSR)
- Initiate application w. customer approval
- Setup account and activate
Bank Branch - Perform underwriting and closing
Office
Figure 4: Onboarding for retail banking application
In line with the definition of a case as stages, Figure 4 shows the retail customer with multiple channels to
interact with the bank, the bank’s CSRs, and the stages of onboarding for a mortgage.
Figure 5: Onboarding for retail banking models
Figure 5 shows the retail mortgage onboarding case as a model with a simplified workflow. At the top of
the figure are the stages for the case with the steps and tasks below. In the case includes a process model
for collecting customer goals for the needs assessment with a user interface (UI) model used in this
process. A decision model is referenced for prioritizing various mortgage loan options for a
recommendation. (Decision models are discussed in more detail in section 2.6.)
Given this example case, we will next look at how low-code models are used to build applications with
layers of specialization.
2.2 Defining an application and grouping models
The Pega Platform models are grouped to define an application, such as customer onboarding. That is, the
case, process, UI, data, decision table, and other models defined for a specific business application can be
grouped for deployment as an application. Figure 6a shows the models from Figure 5 along with others
bundled together into such a group. As models are often called rules in Pega Platform, the group of models
is called a ruleset.
Figure 6: Pega Platform application definition
a. An application as a group of models (a ruleset)
b. An application as layered groups of models (rulesets)
c. An application as layered groups of models (rulesets) with versions
d. Multiple applications from rulesets
Multiple models grouped into a ruleset can be updated, versioned, and deployed together. This ruleset
supports management of an entire application or other combination of models that should be maintained
as a unit.
Often there are enterprise standards for applications, such as standard portal screens, standard corporate
data models, and standard security enforcement. In Pega Platform, these standards are typically included
in the enterprise reuse layer, a ruleset that can be included in the definition of each application to further
enforce the relevant standards. This provides a key ruleset layering technique as illustrated in Figure 6b
where the onboarding rules are layered in a stack on top of both the retail banking standards and
enterprise standards rulesets.
Furthermore, the rulesets can be versioned by three levels of major, minor, and patch/revision 3 versions for
deployment and application evolution. Rulesets are the unit of deployment to a Pega Platform instance.
3
Pega has adopted the Semantic Versioning approach for the Platform where the third set of numbers represents the patch version. However, we leave our clients freedom to decide how they
choose to use Pega Ruleset and Application version numbers. As a result, the third digit is sometimes used as a revision (with features included) and not a patch (bug fixes only).
Figure 6c shows a major application version with a minor version ruleset “layered” on top. An application
can consist of multiple rulesets where the individual rulesets are used to group models by function and by
version. Models in a minor version of the same name as the major version take precedence for execution,
thus facilitating enhancements or patches by just deploying the additional ruleset.
More generally, Pega Platform provides reusability of application components by including the same
ruleset in multiple applications. For example, Figure 6d shows two different applications – the commercial
banking and the retail banking – where the enterprise standards are shared, but the commercial and retail
applications have different onboarding models and additional standards. Pega Platform makes it easy to
share layered rulesets such as the “Retail Standards” application that builds on the “Enterprise Standards”
application. Access control starts at the application level, so a specific user or user group could process
onboarding for both commercial and retail customers by using the Pega Platform applications or be
restricted to processing of onboarding to just one of the applications according to the access control
settings.
The grouping of models into versioned and layered rulesets for applications supports reuse, access control,
and the deployment process. A user (or client API) is given access to one or several applications. A ruleset,
sharable across applications, includes those models that would likely need to be changed together and
deployed together for an incremental update to the application. A pilot deployment for new functionality
or a fix can be created easily by defining a new application with the updated higher-version ruleset and
then giving access to the pilot deployment’s stakeholders.
2.3 Situation-based modeling
Pega Platform has a rich mechanism for creating applications that support many variations of an
enterprise’s business processes. For example, when the customer onboarding process varies by geography
(e.g., state or country), when the loan origination process varies by purpose (e.g., home loan or car loan),
or when the offers available to a customer differ based on the customer’s preferred or non-preferred
status.
We have seen how models can be grouped into rulesets for reuse and deployment as an application and
by the choice of rulesets deployed can create specialized applications. However, the most important way
of creating situational applications is through the use of model classes. When a model is created, it is placed
in a hierarchical name space of classes, similar to classes from object-oriented programming languages in
that the hierarchy defines inheritance but fundamentally different in that the objects of concern are all
forms of models, not procedural programming language objects.
Inheritance means that if a particular named model is referenced, the reference is resolved by looking first
in the current class; if not found, inheritance looks in the parent class; and if not found there, then in the
grandparent class, and so on. And for any model in a parent class, a more specialized model can be created
in a child class.
Classes are most often created automatically, using the tools in App Studio and Dev Studio. For example,
when a case type, such as Retail Onboarding, is defined along with its subcases for specific situations, such
as car loans or home loans, the classes for case models, including associated data properties, are
automatically generated. Figure 7a shows such a class structure.
Figure 7: Specialization of models
Figure 7a. Class Hierarchy of the Retail Onboarding cases
Figure 7b. Class definition with inheritance from two parents
Figure 7c. Circumstancing for fine-grained control of model applicability
This namespace dynamically spans all the models across all the rulesets included in an application. The
inheritance shown is the primary inheritance structure in Pega Platform, but there is a secondary
inheritance as well, which refers to an alternate class for model inheritance. It is secondary because during
dynamic model selection this alternate class is searched after the primary class inheritance. (Primary
inheritance is called “pattern-based” in Pega Platform while secondary inheritance is called “direct.”)
Secondary inheritance is commonly used to enable classes to inherit from the Pega Platform core classes
such as for work items or cases. Work items are the business-meaningful case objects processed in a Pega
Platform application. These built-in classes have a namespace defined by Pega. Secondary inheritance is
also used to inherit from reusable frameworks that have their own separate hierarchical namespace.
For example, as shown in Figure 7b, the FinCo company has a class for company-wide models and, using
primary inheritance, defines specialized classes for each of the US and EMEA divisions. Each division has a
class for its own specialized onboarding models. Secondary inheritance in Figure 7b enables the US and
EMEA models to inherit from a common financial onboarding framework. The framework onboarding class
is itself in the financial framework through primary inheritance, and it inherits from the Pega Platform work
item class through secondary inheritance.
For finer-grained definition of when a model should be used, a set of circumstances can be defined, which
must be true for the model to be used. Such circumstancing is most typically used for situations where just
a few out of a large number of alternative values require unique handling. For example, if just a few US
states have unique handling for a portion of the loan origination process, such as unique compliance
regulations as shown in Figure 7c, circumstancing can be used.
Circumstancing can also be used to specify that certain models have specific effective dates either from a
given date forward or for a specified date range. These date-driven circumstances can be used for
advanced deployment of new models that should take effect at a specific date and time.
At run time, models are selected for execution according to these structures of classes with primary and
secondary inheritance, versioned groups of models (rulesets), and circumstances. The net effect is that
very rich applications can be created with significant reuse and modularity for development and
deployment. (The mechanism for run-time “self-assembly” of the application is explained in more detail in
section 3.2.)
2.4 Decision management and models
Pega Platform decision management uses business rules and predictive analytics to provide automated
decision-making and recommendations such as the next-best-action for a CSR or a customer using a self-
service application. These are key capabilities for customer acquisition, customer retention, cross-
sell/upsell, risk management, and process optimization. From within a process flow, a decision strategy can
be invoked to guide a customer interaction, or predictive analytics can be used to determine the next step
at a decision point in the flow.
Decision strategies are models that define the data used for decision-making, processing of the data that
might be required, the predictive or machine learning models to be used, how the decision is presented to
a user, and in the case of customer offers, a recording of the acceptance or rejection of the offer. All internal
and external data sources available to the application can be used in the decision strategy, including big
data sources and Hadoop processing. The history is used for analytics and can drive self-learning, real-time
adaptive models.
Decision strategies are defined as a visual model showing the data flow and processing leading to the
resulting decision. For example, Figure 5 shows a decision model with the data flow from source to product
recommendation where priority scoring for presenting recommendations to the user is handled by a
decision table. Figure 8 shows the resulting mortgage options presented to the user in priority order.
Figure 8: User screen showing mortgage product recommendations
The types of decision models that can be created in Pega Platform include predictive, machine-learning
adaptive, scorecards, decision trees, and decision tables. For example, Figure 9 shows a portion of the
decision table driving the mortgage recommendation. In addition, models from many other commonly
used predictive modeling products can be imported through the industry-standard exchange format,
PMML.4 The models can be simulated, compared by using historical or real-time data, and monitored with
real-time visualizations of KPIs.
Figure 9: Decision table (partial) for product recommendations
2.5 Declarative computation models
Declarative computation models are a unique aspect of the Pega Platform model-based application
development, allowing non-procedural computation driven by changes in data. This feature provides
significant development simplicity and productivity.
Figure 10 shows the declarative network for the computation of the monthly mortgage (fixed-rate)
payment displayed in the screen shown in Figure 8. The monthly payment output is dependent on the loan
principal amount and two other declarative computations: one converting the loan term in years to the
term in months; and the second, converting the yearly interest rate to a monthly interest rate.
4
It is also possible to call out to other services for model results, including services from Google, AWS,
and others.
Figure 10: Declarative computation for fixed-rate mortgage
By referencing the monthly payment, the screen that is displayed to the user is automatically recomputed
if any of the inputs, such as loan principal amount, have changed since the last computation. This greatly
simplifies the logic of the application, especially for computed values, which are used in multiple models
as well as multiple UI models.
2.6 Design validation

A significant advantage of building applications as models rather than low-level procedural code is that
many errors can be identified at design time, making application testing much easier and reducing errors
in deployed production applications. In Pega Platform, this design validation capability is called a guardrail.
Figure 11 shows the display in Dev Studio with the summary of guardrail warnings in the financial services
onboarding application.
Figure 11: Design validation guardrails
Types of guardrail warnings include data integrity, logic conflicts, performance, maintainability, and
documentation. For example, if a rule in a decision table is unreachable because its conditions for action
can never be met, a guardrail warning will be posted. Or, a caution would be issued if the complexity in a
case design creates so many stages that it would be hard for an end user to understand. Guardrails span
all types of models in an application and have a severity of caution, moderate, or severe in terms of the
potential impact on the application. A weighted guardrail score (0–100) based on warnings, severity, and
justifications is generated for each application. These guardrails provide an overall metric on design
validation.
Section 3: Pega Platform unified, open architecture
and dynamic run time
Pega Platform is implemented as a unified architecture which means that all of the components share
models, operations, and monitoring without requiring custom integrations to tie together dissimilar
technologies.
The Pega Platform engine is at the core of application execution, and it controls the “self-assembly” of
applications. It has a dynamic and efficient approach to selecting, compiling, and executing models to
implement an application. This approach supports productivity and allows for many variations of the
application according to enterprise and customer needs and provides for an efficient and scalable
implementation.
Pega Platform has no run-time model interpreter; rather, it compiles models as needed into standard
execution languages such as Java or JavaScript for superior performance, including the optimization
techniques of the Java Virtual Machine (JVM). A new or changed model is deployed and then compiled when
needed for execution. This is a major difference between Pega Platform and other model-based platforms.
It provides the benefits of low-code development with the scalability and efficiency of an application written
directly in Java or other conventional programming language.
3.1 Overview of the Pega Platform unified architecture and dynamic run time
Figure 12 shows a high-level view of the Pega Platform’s unified architecture. At its center is the Pegaengine,
an instance of which is called a node and is executed in a JVM, often managed within a container. Because
Pega Platform is unified, this applies to all users, whether business application end users or application
developers and administrators. It applies to all types of models whether case, process, UI, decision, or
integration. It also applies to all channels of interaction via APIs.
Figure 12: Pega Platform unified architecture
3.2 Dynamic model selection
Section 2 discussed how modeling can create situational versions of applications by using:
• Models organized into classes to support specialization through inheritance

• Models organized into rulesets that group related models for application deployment and
versioning
• Models with circumstance conditions to handle exceptions and effective dates
These three modeling techniques are the keys to determining how models are dynamically selected atrun
time for execution.
Dynamic model selection takes place as one model references another for execution or retrieval of data.
The resolution of these references through dynamic model selection is a search and prioritization process.
For example, if the retail onboarding case references the retail auto loan subcase for execution, the search
for the proper subcase searches and prioritizes in the following order:
1. According to primary class inheritance

2. According to secondary class inheritance
3. According to ruleset order and version
4. According to circumstance properties
5. According to circumstance effective date and time
Figure 13 shows rulesets defined for the FinCo company with US and EMEA divisions. Ruleset layers are
usually visualized from base capabilities going up to the more specialized, so the inheritance – both primary
and secondary – points downward in this diagram to the more general ruleset when crossing layers. Having
the class structure cross the rulesets is a best practice when the different layers in the class tree are
enhanced, versioned, and deployed separately.
Figure 13: Dynamic model selection with rulesets
Consider the following three examples of dynamic model selection based on Figure 13 that show the use
of classes, rulesets, and circumstances. The numbers in the figures match the example number. The yellow
number shows where the model reference takes place and the gray number shows where the model to be
used is found.
Example 1: Showing class inheritance
The UI for the US Division onboarding case references the image “logo” to be displayed. The dynamic
model selection looks through the data model instances according to primary inheritance and finds an
image with the American flag in the US Retail Division class. It finds no other “logo” defined in the
secondary inheritance the Pega Platform Work Item or in the rulesets in the US application, so the
standard logo for the US Retail Division is selected for use by the UI.
Example 2: Showing ruleset versioning

In an instance of the EMEA division onboarding case, the new customer makes a choice to apply for a
fixed-rate home loan, so dynamic selection looks for the most appropriate fixed-rate home loan
subcase in the EMEA application. Dynamic selection finds such a subcase in the EMEA Onboarding
Ruleset Version 1.0 as well as another of the same name found in Version 1.1 of the ruleset. As the
version priority for the 1.1 ruleset is highest, the subcase from that ruleset will be used. When selected
the first time, the model is compiled and saved in a code cache. For the models that compile into Java,
this cache is directly accessed by the Java ClassLoader.
Example 3: Showing circumstancing
When a New York customer decides to apply for a fixed-rate home loan, dynamic model selection finds
two subcases in the U.S. Onboarding Ruleset Version 1.0 and no others through inheritance. One case
has the circumstance defined for when the “state” property entered by the customer is “NY,” so in this
instance the circumstanced subcase is chosen.
3.3 Dynamic working memory, caching, and persistence
This section explains details of the data management and related optimizations of the Pega Platform
engine execution. Figure 14 shows the Pega Platform engine in more detail.
Figure 14: The Pega Platform Engine
The numbers in Figure 14 are explained below:
1. Dynamic model selection, as explained in the previous section, is optimized through the use of a
model selection cache. For search efficiency, search results are saved in the cache. If a reference to
the same model from the same context occurs again, the result will be found without additional
search.
2. The Pega Platform engine shares dynamic working memory, often called the “clipboard,” among
multiple applications, including multitenant applications. Dynamic working memory is where the
state of applications and client interactions is stored for clients, which can be either a user or a
system interacting via APIs. The figure shows session storage for four clients: the first is a user of
one application, the second is a user of two applications, the third is a user of a multitenant
application, and the fourth is an API client.
The working memory is dynamic in the sense that:

• Data values resulting from a computation can be defined declaratively so that they are
automatically recomputed as source values change.
• Data from external sources, whether direct from a database, middleware, or APIs, can be
automatically refreshed according to usage or schedule, when defined as a Data Page
data model.
3. When selected the first time, the model is compiled and saved in a code cache (and also stored in
the model repository). For the models that compile into Java, this cache is directly accessed by the
Java ClassLoader. The compiled code is cached so that the compilation need not be done again
unless the model changes. The code cache is shared for multiple applications executing the same
models.
Data such as external reference data can be shared across applications in accordance with defined
access controls.
4. The status of core work items is persisted from working memory to the work repository. The work
repository is implemented in a relational database, but with a hybrid storage model where the
properties and state of the work item are stored in a binary format with selected object properties
exposed in relational columns for retrieval and reporting. Pega Business Intelligence Exchange
allows for full extraction of the work items into relational format so that they can be imported into
a business intelligence system for additional analytics and reporting beyond that provided in Pega
Platform.
5.
Session-oriented data goes through passivation and is stored in the work repository according to
configurable parameters, such as a user session being “idle for 30 minutes.” Unreferenced data can
go through passivation separately from the full session, such as data not referenced for 15 minutes.
Activation is done on reference to the data or session, which causes it to be read from the work
repository into working memory.
The optimizations described above allow for a high-performance, scalable implementation of model-based
development and execution by Pega Platform.
Section 4: Pega Platform dynamic user experience
An application with a well-designed UI helps users intuitively navigate and easily understand its purpose.
A well-designed application UI provides the functionality in the right way and at the right time. The problem
is that developing a number of customized apps is a complex and time-consuming task, and many
businesses are still impeded by disjointed, complicated development environments and a lack of trained
designer resources.
Pega provides a simple, low-code platform that empowers the team to quickly build applications with rich
cross-channel UIs tailored to changing needs. This feature enables business users and developers to
collaborate and easily build robust apps that improve operational efficiencies, look great, and are
accessible anywhere.
Pega Cosmos design system (Cosmos) leverages 30 years of best practices to present the application in the
most intuitive way possible. The system is created for and around artificial intelligence (AI), case
management, and collaboration, and is tested against enterprise-specific metrics. The Cosmos user
experience (UX) framework is built to meet the needs of three different user groups: developers building
low-code applications, employees managing daily workflows, and directly engaging customers in
consumer-facing applications.
The designer prioritizes the elements that need to be shown, and Cosmos determines the best way to
present them. These templated best practices deliver intuitive and easy-to-use applications for any end
user, employee, or customer. Training costs can be reduced or even eliminated. Errors are also dramatically
reduced, further reducing costs.
Cosmos aligns with Pega’s low-code design strategy and supports rich, dynamic, and situation-based UX
across all devices. The use of common models allows for an omnichannel user experience supporting all
interaction channels including mobile, web, embedded web portals or mobile mashups, social channels,
correspondence (email and print), and telephony.
Case management – and the storage of case objects in Pega Platform – ensures a unified cross-platform
experience, capabilities, analytics, and content across all the interaction channels. A user can start an
interaction in one channel and switch to another midway through the process. For example, a customer
can start product selection on a mobile device and then complete it on the desktop or via a phone call to
the call center. Pega Platform maintains the context of the interactions across all these channels and
devices in the case object for both the customer and the CSR.
Support for standard look and feel, branding, and screen layout by using skins in App Studio significantly
helps UX design and consistency. Skins enable a common appearance for all application screens on any
device with no specific screen-by-screen setup required. Customizations, such as different branding for
each business unit in an enterprise, are easily specified.
4.1 Dynamic user experience
Pega Platform enables the design and development of a dynamic customer-centric UX across any device,
and it does so without requiring the UX designer or developer to write any code. Pega provides a rich set
of low-code tools for designing, developing, and deploying a UX that works across channels.
Pega’s open, responsive, and adaptive UI technology makes multichannel deployment quick and hassle-
free. Developers can build seamless mobile apps completely in Pega Platform, enhance existing interfaces,
or embed Pega Platform into existing apps or web pages by using the DX API or mashups.
Figure 15: Responsive user interface – example
Furthermore, the UX can be responsive, adaptive, or contextual. Each of these is explained below:
• Responsive: Pega Platform models can specify a responsive user interface that automatically
adapts to the screen size, client platform, and orientation. This is done in Pega Platform with no
programming. Developers can define what happens for ranges of screen sizes. The required UI
code for the mobile client is generated.
For example, Figures 15a, 15b, and 15c show a commercial real estate lending process. Figure
15a shows a responsive layout when using a desktop display. Note the navigation pane is on the
left, with the process steps and opportunity insights taking up the rest of the screen. Figure 15b
shows the layout for a tablet screen with the vertical ellipsis menu placed in the upper-left corner
and opportunity insights below the process steps. Figure 15c shows how the UI is displayed on
a phone, with content formatted for navigation through the sliding menu.
Figure 15a. Desktop screen for commercial real estate lending process
Figure 15b. Tablet screen for commercial real estate lending process
Figure 15c. Phone screen for commercial real estate lending process
• Adaptive: While the responsive UI is based on the client code adapting its content to the screen
size, the adaptive UX leverages modeling to customize the content appearing in the UI according to
the device being used. For example, a desktop user might have access to all options for creating
and advancing a case, while a phone user might have access to only approvals and the most
commonly used options for advancing a case. Circumstancing (as described in section 3.2 Dynamic
model selection) applied to the user interface models can be used when the user experience needs
to account for the differing usage across devices.
• Contextual: A contextual UX is specialized according to the attributes of the current situation such
as the user role or type of case being processed. For example, in the case of applying for a mortgage,
certain documents, including the formal loan application, need to be uploaded to the bank. Figure
16a shows a customer’s self-service screen for uploading documents, while Figure 16b shows the
view of this step for the bank’s CSR.
Both screens represent the same step in the internal process flow, but different user interface
models are used for the customer versus the CSR role. Both are allowed to upload documents, but
only the CSR is allowed to change the document status, for example, from “Pending in Good Order
Review” to “In Good Order.”
Figure 16a: Customer self-service interface
All of these approaches to providing a dynamic user experience – responsive, adaptive, and contextual –
are used to deliver a customer-centric experience according to the usage intent and interaction device of
the user.
Figure 16b: CSR interface
4.2 Mobile apps
Pega Platform provides advanced capabilities for the development of model-based mobile apps without
separate programming efforts or special deployment skills. The framework empowers teams to more
quickly build sophisticated apps that look impressive out of the box because of their responsive and
adaptive user interface. This functionality is enabled through Cosmos as well. The app is rendered as
needed based on the device.
Pega Platform supports both mobile responsive web apps (access through a mobile browser) and native
mobile apps. The mobile web capability leverages the responsive user interface, described in section 4.1,
so that the app automatically adapts to the screen sizes of smartphones and tablets.
a) Figure 17a illustrates the mobile web option where responsive web code, that is, HTML5, JavaScript,
and CSS, is downloaded for a mobile browser session. Apart from configuring the dynamic layout
model, no changes are required for Pega Platform applications to operate in a mobile web
environment.
b) The hybrid native application option in Figure 17b leverages both web and native controls to provide
a rich mobile experience. The container, called Pega Mobile Client™, provides APIs for accessing
device capabilities as well as security services, including user authentication and encryption. Native
capabilities that are supported include geolocation, camera, push notifications, QR/barcode
scanning, and local storage. A hybrid native app for a single Pega application can be built by selecting
the appropriate option in App Studio and then providing the logo and branding to be included in
the distributed app. The app can be deployed through the Apple App Store, Google Play, or any
private enterprise app store. The hybrid application can be deployed by using either stores or third-
party mobile device management (MDM) solutions, for example, Microsoft Intune app manager.
c) Often an enterprise builds an application in the native OS programming languages, Java for Android
or Objective-C (or Swift) for iOS. For example, think about a retail bank that might have a widely
deployed consumer application built this way; or imagine booking a hotel room, upgrading a service,
reporting a problem, or purchasing tickets – all of which can be fulfilled by Pega apps that are
embedded in existing apps through mashup. To make inserting apps seamless, Pega Platform
provides a mashup and an API SDK for iOS and Android, which you can use to embed a Pega
appwithin a native application as shown in Figure 17c. When Pega Platform powers the back-end
process, you can insert that process into an existing app to enable customer self-service. This
capability can greatly enhance the value of consumer-facing (B2C) apps or broaden and deepen the
capability of enterprise (B2E) apps.
Figure 17: Options for mobile application deployment
Pega Platform allows for low-code development of mobile-specific capabilities, including push notifications
and offline mobile.
4.3 Mobile offline support
Apps built in Pega Platform can be quickly configured to run in disconnected mode on a mobile device.
While offline, users can still use key functions, such as:
• The ability to create and complete work.

• The ability to add photo attachments.
• Support for subflows and paragraph rules, as well as the ability to go back to a previous step in
a flow.
• Support for multilanguage localization, which provides the ability to localize an offline app for a
specific language. This support enables the storage of multiple languages and the ability to switch
languages mid-session while offline.
• Support for parameterized data page querying and filtering.
• Support for large data page efficiency and performance.
• Support for data synchronization intervals configuration, which gives application developers more
control over the data synchronization process.
• The ability to run Pega Platform offline-enabled apps through Pega Mobile Client for Windows, by
using a dedicated Windows application instead of a web browser.
4.4 Intelligent Virtual Assistant
Pega Intelligent Virtual Assistant™ (IVA) is a Pega Platform capability that enables the extension of any Pega
enterprise application to have a simple, conversational user experience (UX). The feature provides self-
service or preemptive service by easily turning any voice or text-based channel into an AI-powered virtual
assistant, using natural language processing (NLP) and text analytics to deliver personalized engagement.
Pega IVA is multi-channel ready, and provides a visual interface to configure, manage, and test how the
assistant performs on various channels, and to effortlessly deploy updates in real time, without downtime.
Pega IVA can decipher intent and context to enable better and more intuitive interaction by using a more
natural and casual tone and vocabulary, and perform entity extraction to pull structured fields out of
unstructured text.
Figure 18a. Easy conversation configuration and management across channels
Figure 18b. Integrated conversation simulator for modifications performance tests
c. Real-time and no-downtime updates deployment
4.5 Email bot
Email is still a heavily used channel in both B2B and B2C service scenarios. However, without the
appropriate strategy, processing emails is extremely time-consuming and sluggish as email
communication is inherently unstructured, attachment-heavy, and disconnected from back-end systems.
Pega Email Bot™ allows organizations to rapidly triage and process massive volumes of incoming service
emails to accelerate customer responses, eliminate manual work, and streamline operations.
Pega Email Bot uses easy-to-configure, easy-to-train AI to understand email. NLP text analysis models
automatically extract topics, sentiments, and business entities (such as location and account number) from
emails across multiple languages. Simple intelligent routing rules use AI analysis to automatically reply to
customers in context, triage work to the right person or team, and automate customer outcomes with
Pega’s case management and low-code development tools.
For processes that still require human interaction, Pega Email Bot comes with an AI-assisted email
processing toolbox called Email Manager. Email Manager uses the results of the NLP text analysis to guide
agents to faster resolutions by providing recommended actions and responses.
Lastly, processing email is important, but it is critical that the work generated through email is captured,
tracked, and resolved. Case management within Pega Platform ensures that email is not just answered,
but that the work associated with email is resolved completely, driving to a business outcome.
4.6 Integrated social collaboration
Pega Platform includes an integrated social collaboration capability called Pulse, which adds a social activity
stream to the Case Management user interface. This feature can be made available to every user of all
Pega applications to facilitate collaboration and conversation among application developers, enterprise
users, or customers. You enable the feature by including the Pulse ruleset in the application definition.
By unifying the user experience across departments, Pulse helps drive productivity with a customized,
collaborative dashboard. With all Pega applications united across sales, marketing, and customer service,
Pulse enhances collaboration opportunities among:
• Customer service personnel and product managers to resolve customer issues.

• Account executives on a sales opportunity.
• Marketing personnel concerning campaign design.
• Insurance underwriters making new insurance policy decisions.
• Financial services professionals approving a loan or credit card request.
Pulse and other use cases that Pega Platform provides offer customers a robust notification framework
that, among others, includes email digests, replies through email and push messages, and notification
preferences. Furthermore, users can share instant messages, files, and URLs with others in their work
group. These useful capabilities include:
• Contextual activity streams providing public or private posts in the context of a case or user.
The stream scope depends on where the post is used in the interface. For example, when the
activity stream is embedded in a case user form, only posts that are related to the case appear.
Users can choose to post information privately for users working on a case rather than
broadcast to the entire work group.
• Other data sources that users can now include in their feed to make them more robust. This
specific capability is highly used in the Pega Sales Automation™ application.
• User profiles including operator information such as photo, position, contact information, and
skills. Users can modify their own photos and biographical information.
• Actionable posts where users who have the correct privileges can open cases or create tasks
directly from a post. Pulse can be used directly in workflows. For example, messages can be
posted when a case reaches a deadline or goal.
• Own post types that users can now create and add to extend the list of three standard types
that are set as default in Pega Platform, that is, standard post, private post, and task. Log Activity
and Close Plan are examples of post types in Pega Sales Automation, while other basic examples
include Poll (survey), Check In (location), or GIPHY.
Figure 19: Pega Pulse – Integrated Social Capabilities
4.7 The Digital Experience (DX) API
For organizations that want to plug Pega process, case, and decision logic into other front-ends, Pega offers
the Digital Experience API. The solution enables front-end developers to create their own front-end
interfaces across all digital channels while directly connecting them to end-to-end processes that help
organize operations within the enterprise.
With the DX API, developers can not only jumpstart interface design with pre-built templates but also gain
the flexibility to utilize a number of useful and difference-making capabilities. They can use the UI
framework of their choice, such as React and Angular, and capitalize on Pega's powerful UX design system
to create connected customer experiences with their preferred tools. Because the DX API provides dynamic
meta-data to the front-end layer, changes in process steps or data models can be dynamically reflected in
the front-end without recoding.
The DX API exposes Pega Platform capabilities as a REST-enabled service to power the organizations front-
end UI framework of choice. Pega Platform provides a rich set of APIs that not only deliver the data for UIs,
but also populate the UI metadata to layout screens. This functionality enables developers to make UI
changes (for example, one- to two-column layout) in Pega Platform, and then have them reflected
automatically in the application.
4.8 Pega Process Extender for Salesforce Lightning
Organizations that use Salesforce can also greatly benefit from Pega DX API to manage Pega Platform cases
within their existing Salesforce solution through Pega Process Extender for Salesforce Lightning – a bundle
of web components that run Pega Platform cases natively inside Salesforce.
Pega Process Extender for Salesforce Lightning provides world-class intelligent digital process automation
(DPA) inside the Salesforce implementation. Pega Platform cases are rendered in native Salesforce
Lightning Design System (SLDS) leveraging the DX API. Changes to screen layouts or processes that you
make in Pega Platform are instantly reflected in the Lightning screens without recoding.
With the Pega Process Extender for Salesforce Lightning, organizations can use their existing Salesforce
user interface with Pega Platform dynamic case management to manage and automate work and
processes from end to end. The key features include:
• No need to style screens because Pega Platform cases are rendered in native Salesforce Lightning
Design System (SLDS)
• No need to download JavaScript/CSS
• Dynamic/real-time refresh
• Automatic authentication and single sign-on
• Drag-and-drop installation
• No coding required
For examples and starter packs for VUE, React, Angular, and Pega Process Extender for Salesforce
Lightning, go to Pega Marketplace.
4.9 Pega Web Mashup
Pega Platform also provides capabilities that embed all or some Pega applications inside of existing
applications, such as web and mobile apps.
Pega Web Mashup enables a Pega application to be embedded as a gadget on the pages of a web
application located on an intranet, extranet, or internet site regardless of the underlying technology that
the web application was developed in. Pega Web Mashup can utilize existing authentication and
authorization patterns to enable a seamless login process, and provides a robust JavaScript API to complete
such tasks as passing information back and forth between applications and subscribing to Pega Platform
page events, for example onLoad or onConfirm.
Pega Mobile Mashup provides similar functionality to Pega Web Mashup for existing native mobile apps.
Section 5: Pega integration capabilities
Enterprise business applications do not exist in a vacuum. They must communicate with other systems
and services in the organization, and often partner systems as well. Pega has a comprehensive suite of
integration capabilities that allow Pega applications to fit seamlessly into an enterprise ecosystem. Pega’s
approach is to be open and ensure that we are leveraging best-in-class third-party capabilities while
providing end users with a unified experience through our unified platform. Pega Platform integration
capabilities are broken down into the following categories:
• Data Integration
• Live Data
• Pega Robotic Process Automation
• Digital Experience (DX) API
• Web Mashup
5.1 Pega® Data Integration
All the Pega Platform integration options are forms-based, so that no custom code is required to connect
to an external system. Whenever possible – when metadata formats such as WSDL or Java introspection
exist – Pega Platformconsumes services or publish services automatically. The Pega Platform technology
can be declaratively invoked when needed by simply referencing the data managed by the adapter.
Figure 20: Supported integration technologies
Note: technologies listed in blue are not supported in Pega Cloud Services deployments.
The Pega Platform data integration capability is broken down into four areas:
• Connectors: These are outbound connections, in which Pega Platform is invoking a service or calling
a function that is provided by an application. Pega Platform supports connectors for multiple
services. These include web services APIs through REST, SOAP, MQ, or HTTP, and the ability to
invoke robots, part of Pega RPA (see more below). If metadata exists for the services that Pega
Platform is invoking, such as a WSDL file for a SOAP connector, Pega Platform provides wizards that
can consume the metadata and generate data access objects, bindings, and so on, automatically.
• Pega API: Provides a standard set of operations that deliver API support in the areas of application,
DevOps, user management, data privacy, digital experience, and system administration. These built-
in REST/JSON services enable a rapid implementation of mobile and client applications. By default,
Pega API services are secured with user credentials and TLS/SSL connectivity.
• Services: These are inbound connections, in which Pega Platform exposes some functionality for
another engine or application to invoke. Pega Platform can receive requests via SOAP, MQ, HTTP,
or RESTful services. If the service type has associated metadata associated, Pega Platform can
automatically generate it, for example, by publishing the various steps of a business process as a
WSDL.
• Data Sets: Pega Platform also includes a whole collection of models – data sets – that are primarily
geared toward event stream processing and big data support. The capabilities provided here enable
applications that can take in high volume event streams directly, consume streamed event data
from distributed messaging systems like Kafka, or pull data records from Cassandra- and Hadoop-
based systems to support real-time decision management.
The Pega Platform data integration capabilities can be used to integrate directly with the existing systems
or can use the enterprise service bus (ESB) or service-oriented architecture (SOA) infrastructure. Pega is
focused on solving the middle-out problem of automating and improving business processes, not the
bottom-up, infrastructure challenges that SOA and ESB systems address, or the top-down approach where
only one communication channel is optimized.
5.2 Pega Live Data
Pega Platform has a built-in data virtualization capability called Pega Live Data that simplifies the ability to
dynamically deliver the right information to cases and screens.
Figure 21: Pega Live Data Integration
Pega Live Data provides a data abstraction or virtualization layer. Users who are configuring a Pega
Platform application are simply working with the data model and are not concerned with the origin of the
data. Abstraction minimizes confusion and accelerates project timelines by showing just what is needed,
which enables organizations to take a more pragmatic approach to infrastructure changes. Finally,
abstraction supports data from very different sources, like cloud and mainframe, to be used to drive a
business outcome. Organizations are able to derive value from all available data, regardless of the age of
the data.
Live Data is declarative in nature. Citizen or professional developers do not have to explicitly instantiate or
call interfaces to populate data in objects; they simply refer the model and properties and Live Data
determines whether the data needs to be fetched or refreshed.
Live Data lets you easily configure conditional access, for example, in situations where the data model
might be populated from different systems of record based on a condition or business rule. Live Data can
do meditation at run time to determine which system of record should be used to populate the data model.
In situations where the data model is sourced from multiple systems, Live Data provides support for
composite data sources. For example, if there is a Customer data object, general customer information,
such as name and address, may come from the CRM system that the business uses. Customer billing
information might be housed in the billing system. It might make sense to join that information in
populating the Customer data model. Live Data can populate that data model from the CRM and billing
systems. Furthermore, that data can be retrieved in parallel so that performance service-level agreements
(SLAs) are supported.
Live Data also provides robust caching support, native in Pega Platform. In situations when real-time data
is not required, for example, when a list of products is in use, Live Data can cache the information in
memory. Caching can be at session scope all the way to cluster-wide scope. In addition, Live Data
determines how long information stays in cache – based on a business rule or a defined time interval.
5.3 Pega Robotic Process Automation

Organizations are under tremendous pressure to meet the rising service expectations of digital customers.
While customer experience continues to rise to the top of every business agenda, customer service
representatives (CSR) struggle to navigate a combination of disconnected desktop applications whenever
they engage with customers. Distracted employees waste time toggling between dozens of different
applications to input or review customer data, frustrating employees and customers alike.
Figure 22: Pega Robotic Process Automation
Pega has rich support for both attended and unattended robotic process automation (RPA). Attended
robotic process automation (or robotic desktop automation) helps employees work faster and more
efficiently, simplifying and automating business processes and transactions. RPA accelerates productivity
by optimizing the way work is performed. These “personal robots” run collaboratively on every CSR’s
desktop to automate key tasks and workflows. That way, employees can deliver an exceptional service
experience by focusing on what matters most – the customer.
With unattended RPA, organizations can automate the mundane, tedious, time-consuming, manual work
that is hampering productivity and efficiency. Whether adjudicating claims, onboarding customers or
employees, reconciling financials, updating customer information in the system of record, Pega Robotic
Process Automation can manage the back-office work across the enterprise.
With Pega Attended RPA™, the robot (Robot Runtime) is typically started by an end user. When the robot
starts, it loads a solution package that contains automations. The robot then watches for specific events
that are defined in the solution. These events signal an automation to start. The event that starts an
automation can come from a Pega application or an any other application that is running on the
workstation, as long as the application is specified in the automation. When deployed as an unattended
robot, Pega Robot Manager assigns work to the robots similar to the way humans are assigned work within
Pega Platform.
Pega’s RPA solution comprises a number of integration components including brokers, translators, natural
language processors, adapters, drivers, and connectors. To create and integrate their automations,
developers use Pega Robot Studio. Robot Studio organizes adapters and robotic automations and provides
all the features of a modern development environment.
Pega Robotic Automation can be combined with Live Data to access legacy systems that lack modern APIs.
If APIs become available, simply swap out the data source by using Live Data.
Figure 23: Pega Robotic Automation with Live Data integration
Section 6: Pega Platform and artificial intelligence
Pega Platform and Pega Customer Decision Hub™ (CDH) leverages AI technologies such as predictive
analytics and adaptive learning (aka machine learning), natural language processing (NLP), and complex
event processing to:
• Determine customer context: Interpret the signals generated by each individual to determine
whether the right time to sell, serve, retain, or just reduce risk by taking no action at all.
• Calculate propensities: Analyze all the possible offers and actions, in real time, while the customer
is interacting to determine exactly which messages are relevant.
• Project the value: Calculate financial value in real time, for each unique customer and situation, to
determine what is best for them and the business.
• Execute across channels: Engage the customer, using the channels and treatments that are most
relevant to them. This is made possible thanks to AI, which can tell when a customer is in-market,
triggering a marketing touch.
Figure 24: Pega Customer Decision Hub – Component Architecture
Pega’s predictive analytics leverage real-time and historic big data to calculate likely customer behavior,
which can be easily embedded into action strategies, enriching decisions and improving outcomes. Real-
time decision management automatically executes the next best action that aligns with the customer, the
specific interaction, and company objectives. )
Adaptive analytics is a form of predictive analytics that leverages self-learning adaptive models to improve
predictions about how customer interests and needs change. Adaptive analytics captures and analyzes
data to deliver predictions in situations where historical information is not available. By updating the
models after each response to a proposition, Pega Platform increases the accuracy of decisions. Adaptive
decision management identifies propositions that customers are most likely to accept, improves customer
acceptance rates, and balances customer needs with business objectives.
Pega Platform includes natural language processing (NLP) capabilities to process and structure text data:
• Categorization models that assign incoming text to a predefined category:

• Sentiment analysis: Detect and analyze the feelings (attitudes, emotions, opinions) that
characterize a unit of text.
• Topic detection: Assign one or more classes or categories to a text sample to make that text
easier to manage and sort.
• Intent analysis: Determine a user's intent in social media posts, comments, messages,
emails, and so on, to find out whether that user is likely to subscribe to services or buy
products.
• Text extraction models that extract named entities from text data and assign them to predefined
categories, such as names of organizations, locations, people, quantities, or values:
• Rule-based text extraction: Extract entities that follow a specific text pattern, for example,
dates, account numbers, emails, and so on. Use Apache RUTA scripts to build rules for text
extraction.
• Model-based text extraction: Extract text with a supervised machine-learning model.
• Keywords-based text extraction: Identify a set of keywords that pertain to the use case,
for example, different names of the same product.
• Auto-tags: Dynamically identify the phrases that capture the essence of the text.
Pega Event Strategy Manager (ESM) enables organizations to listen to all types of customer signals, find
meaningful patterns, proactively anticipate customer needs, and act on opportunities immediately –
turning streaming data into revenue streams. Examples include:
Figure 25: Pega Event Strategy Manager
• Web and back-end transaction systems that record the products that customers buy or are
interested in.
• Mobile devices that indicate where customers are located and where they might be going.
• Intelligent devices connected to the Internet of Things that can determine whether and
how urgently customers (or their equipment) need help.
• Social network feeds that hint whether their relationship with the organization is improving
and how much influence they are having on others.
Section 7: Cloud at Pega
7.1 Cloud Choice
Pega’s Cloud Choice Guarantee program is designed to meet the varied needs of our clients all around the
world. This program assures clients that they are not technically locked in to any cloud infrastructure
provider, while providing the flexibility to manage their own architecture on a certified infrastructure
provider. Our guarantee provides the assurance that we will continue to certify existing platforms to run
Pega Platform, and should the client elect to migrate between approved clouds, they can do so without
license penalty. Most importantly, the Cloud Choice Guarantee allows clients to reduce their operational
burden by migrating to Pega Cloud at any time.
Figure 26: Pega Cloud Services Scope
7.2 Pega Cloud Services
Pega Cloud Services offers the fastest path to value with Pega Platform. With Pega Cloud Services, we
provide you with the environments, tools, and services to get you up and running, fast. In fact,
environments are typically available within 5 business days, allowing clients to focus on developing
applications, rather than managing cloud environments.
Pega Cloud Services delivers and ensures a secure, compliant infrastructure, freeing teams to focus on the
value that their Pega applications deliver, not the infrastructure that runs them. Pega Cloud Services
alleviates the operational burden on organizations, enabling them to develop, configure, deploy, manage,
update, or change the Pega applications that support their business. Pega Cloud Services:
• Configures and provisions instances.

• Automates deployments, backup, failover, cache tuning, and more.
• Enables operation management including monitoring and response, patching, upgrades, and
performance troubleshooting.
• Provides security services and tools including encryption, whitelisting, encryption (in transit and at
rest), antivirus/malware configuration.
• Includes infrastructure and architecture to support Pega Platform.
• Manages ongoing upgrades to ensure that the latest, greatest version of Pega Platform is being
leveraged in the organization.
A Pega Cloud Services subscription includes all of the environments the team needs to continuously
improve Pega applications. Dev/test environments are configured to support branch development and unit
testing. Staging environments support scale benchmarking. Finally, Production environments are built to
support licensed and configured Pega solutions.
Pega clients that decide to capitalize on Pega Cloud Services have several options to choose from. The
standard subscription with the base package can be extended to include various add-on options:
Standard production
Cloud only options Test/dev only options
package
Complete Pega Cloud-based
Standard dev/test sandbox Standard dev/test sandbox systems for development and
testing
Large dev/test/pre- Large dev/test/pre-
Base package
production sandbox production sandbox
Pega Cloud production Production mirror sandbox
Additional storage
environment subscription
Additional cloud data
Bundled production storage
storage
Additional cloud file
Pega® Diagnostic Cloud
storage
VPN connectivity
Pega Cloud HIPAA Edition
7.3 Client-managed cloud
For organizations that have already invested in another cloud architecture or that use several different
cloud infrastructures, Pega also offers a client-managed cloud. In this model, organizations deploy and
manage Pega Platform using their own teams on Amazon Web Services, Microsoft Azure, Google Cloud
Platform, VMware Tanzu, or RedHat OpenShift.
For clients managing their cloud environments on supported infrastructures, Pega Platform 8.2 and later
are optimized for orchestration using Kubernetes. Pega software’s architectural shift to Docker decouples
Pega software from the underlying infrastructure, in order to deploy Pega applications in orchestrations
sized to your business needs. Leverage the flexibility of Kubernetes to reduce your operational burdens
and optimize Pega workloads in both your Pega applications, such as the suite of CRM applications, and in
your Pega application extensions.
Section 8: Pega Platform enterprise-class applications
Previous sections have explained how Pega’s approach to low-code development facilitates productivity,
specialization of applications to a wide variety of users and situations, and a dynamic user experience that
includes broad mobile support. Because Pega Platform is intended for mission-critical, strategic
applications in an enterprise, the applications must also have enterprise-class quality of service. This means
that the applications are secure, easy to deploy and manage whether on-premises or in the cloud, are
scalable and reliable, and integrate well with the existing enterprise infrastructure and SOR.
The following sections provide details about Pega’s enterprise-class security, deployment and
manageability in the cloud and on-premises, as well as its reliability, scalability, and integration.
8.1 Enterprise-class security
Pega’s low-code application development approach provides a key advantage in building secure
applications. Since the code is generated from models with embedded security best practices, the code is
secure and eliminates the kinds of errors introduced in conventional programming. Pega-generated code
uses best practices to protect against key application security risks 5 such as injection, broken
authentication and session management, and cross-site scripting.
Pega Platform supports fine-grained authorization, including role-based access control (RBAC) and
attribute-based access control (ABAC), which are defined on the objects and functions of the application
model. Full capabilities for authentication – including use of LDAP, SAML SSO, Multi-factor, and Active
Directory – as well as multiple authentication schemes, are supported at the same time. This supports
taking differing approaches for internal users, partners, and customers. Models (rulesets) can also be
locked and passwordprotected to restrict changes.
Pega Platform provides powerful capabilities for ensuring and implementing security in applications,
especially when deploying guardrail-compliant applications. In most cases, Pega Platform’s model-driven
architecture allows for applications to be secured by configuring built-in features, without the need to rely
on custom code that might pose security risks due to the developers not being security experts.
Pega’s enterprise-class security helps enterprises ensure that the three major pillars of security–
confidentiality, integrity, and availability– are maintained at the highest level. This is primarily accomplished
through authentication, authorization, and auditing.
All of the capabilities necessary for an application to be compliant with various security standards and
regulations, such as the Payment Card Industry Data Security Standard (PCI DSS), and the Health
Insurance Portability & Accountability Act Security Rule (HIPAA), are provided. Audit trails are provided
through normal user activity history logs, as well as specific security alerts for situations such as users
trying to access restricted activities, data, or reports.
5 OWASP top 10 security vulnerability risks: https://www.owasp.org/index.php/Top_10_2013-Top_10
In addition to features that explicitly enforce authentication, authorization, and auditing, other Pega
Platform components represent important policies, assets, and safeguards for concomitant use, such as:
• Certificate, key, and token management to secure the functioning of other security features.
• Data confidentiality and encryption to secure sensitive information at any point in a business
process.
• Virus checking to check programs before processing any email or attachment.
Content security policies to lock down applications and mitigate the risk of unwanted content.
For more information, see the Guide to Pega® Cloud security and Security basics.
8.2 Enterprise-class deployment and manageability in the cloud or on-premises
Low-code application development with Pega Platform provides key benefits in application deployment.
When migrating from development to QA to production, users do not move any code– instead, the models
are moved as versioned rulesets. In this way, incremental updates can be deployed for incremental
functions or fixes. Multiple versions can be deployed simultaneously to support pilot testing and then
migration, all without system downtime. Users are directed to their proper version according to their
application access settings.
The Pega Platform engine and the code generated from business models includes embedded
instrumentation and notifications, so that administrators can assess and monitor the performance and
integrity of their Pega-based solutions.
Sitting above existing APM tools and supporting both on-premises and cloud-based systems, Pega
Predictive Diagnostic Cloud™ (PDC) monitors the Pega application layer using a unique knowledgebase.
Pega Platform sends PDC application-agnostic health pulses, and performance and usage data. In addition,
on encountering performance, integrity or stability events, Pega Platform asynchronously sanitizes and
sends an alert message to PDC. The PDC service then performs identification, correlation, aggregation,
prioritization and notification services as appropriate, based on the content of each and every
message. PDC provides senior stakeholders with information about application usage and health, guides
technical managers by identifying the application changes that will provide the greatest improvement, and
improves developer productivity by providing interpretation, aggregation and navigation to aid in their
investigation and remediation. PDC also simplifies the support process by sending notifications for urgent
events, consolidating resource information, and providing an easy to use tool for ad-hoc investigation and
trouble shooting.
Figure 27: Pega Predictive Diagnostic Cloud (PDC)
In addition, authorized users can use Admin Studio to gain insight into current users, requestors,
background processes, memory utilization, and more, within Pega applications.
8.3 Enterprise-class reliability and scalability
Pega Platform is designed to operate on a large scale with a single cluster supporting tens of thousands of
simultaneous users and millions of cases per day. Both horizontal and vertical scaling is supported:
• Horizontal scaling with near-linear scalability within a certain range of scaling can be supported by
deploying Pega JVM nodes across multiple load-balanced servers. A shared database architecture is
used across a horizontal cluster, with certain instances of communication between the nodes taking
place throughout the database. Hazelcast is generally used as a clustering solution for the nodes. If
needed, database scalability is provided by the database vendor’s clustering capabilities.
• Vertical scaling is supported by running multiple Pega JVM nodes on a multi-CPU server.
Pega Platform adds capabilities for failover in the case of node server crashes. Browser crashes are also
handled by recovering the user session.
8.4 DevOps and automated testing
Application evolution is facilitated by Pega’s DevOps methodology. This is a set of tools and best practices
that address issues in three crucial spaces: development, operations, and quality. Pega Platform offers the
right tools to support DevOps, including standards-hooks into most popular testing and pipeline
management tools. Pega Platform’s native DevOps capabilities make open, automated testing easy, even
for organizations that don’t have DevOps pipelines in place. With no-code testing built into Pega’s CI/CD
(Continuous Integration/Continuous Deployment) pipeline, as well as support for open tools, organizations
gain ultimate flexibility in guaranteeing their applications’ quality.
Deployment Manager, which is also part of Pega Platform, can be used to manage the end-to-end
application development pipeline in support of DevOps requirements and initiatives. However, if Jenkins,
Artifactory, or another tool is already in place, Pega Platform’s open architecture makes it easy to simply
incorporate these tools and processes into the existing DevOps framework.
Using the native case management and process capabilities of Pega Platform, Pega® Deployment Manager
enables users to define CI/CD pipelines that they can use to swiftly move applications through consecutive
stages, from development to testing to deployment. While Pega Platform’s model-based structure makes
development uncomplicated, it is testing and deployment that underline the evolution of application
building with Pega.
Deployment Manager allows for convenient, one-click deployment by configuring and running CI/CD
application workflows from within Pega Platform, as well as through full automation of workflows such as
branch merging, application package generation, artifact management, and package promotion to
different stages in the workflow. Ultimately, this eliminates the need for third-party tools and ensures a
standardized and predictable deployment process, resulting in high-quality releases and faster time to
market.
Pega Platform provides enhanced application testing functionalities to enable more comprehensive and
flexible tests through the use of both built-in tools and third-party solutions. Some of the enhancements
include the new, aggregated application-level test coverage report, REST API-based start and stop coverage
sessions, customizable quality-related settings on new landing pages, and the additional quality metrics
that are available on the Application Quality dashboard.
Figure 28: The Application Quality dashboard
Section 9: Pega Platform — digital transformation
Pega Platform is designed for building, deploying, and evolving business applications that embody the
uniquely differentiated activities of every enterprise. This technical brief has shown how Pega Platform is
a unified, open platform that provides state-of-the-art enterprise application agility through the following
key capabilities:
• Full low-code application development providing outstanding productivity, empowering citizen

developers, and bringing agility to typically difficult-to-implement capabilities such as integration
and predictive analytics-based decision management.
• Dynamic runtime self-assembly of applications enabling implementations unique to the specialized

needs of customers, partners, and enterprise users, according to factors such as customer status,
product type, and geography.
• Dynamic, omnichannel user experience that includes social collaboration for all web, mobile web,
and native mobile apps based on the same models.
• Complete enterprise-class applications that are secure and easy to deploy and manage, both on-
premises or in the cloud, and that are scalable and reliable, and integrate well with existing
infrastructure and systems.
These capabilities are all part of a unified platform in which applications can be built using an entirely
model-based process, with no conventional programming required, while also achieving enterprise-class
performance and scalability.
9.1 The broader capabilities of Pega’s applications

Beyond the core Pega Platform capabilities, Pega also provides packaged applications built using Pega
Platform that deliver strategic horizontal and vertical business capabilities, while also retaining the
flexibility to customize the application to meet an enterprise’s specific business needs. Each application is
delivered as a set of models.
Horizontal applications, such as Pega® Customer Service, Pega Sales Automation, and Pega Customer
Decision Hub, provide business solutions across industries. Vertical applications provide business solutions
for specific industries including financial services, insurance, healthcare, telecommunications, government,
and manufacturing.
The example used in previous sections of a financial services onboarding solution is based on Pega
Onboarding for Retail Banking. As shown in Figure 22, this application is based on the Pega ® Financial
Services Foundation, which is itself based on Pega Platform.
F
igure 29: A typical Pega application stack
With Pega’s Horizontal and Vertical applications, an enterprise can quickly develop their own
customizations in the form of modifications and extensions that are simply layered on top of the base
application.
Conclusion
Many organizations looking to start their digital transformation journey begin in the wrong place, either
with a pricey tech refresh, or with a siloed, channel-centric application. Pega’s approach allows
organizations to start “middle-out” with a microjourney – a specific business process that drives a
meaningful outcome. Organizations can then simplify and automate this process using Pega Platform’s
low-code application development methodology, easily connecting it to both new and existing channels
and user interfaces. And with Pega Platform’s data abstraction layer, the information needed to complete
the microjourney can be pulled in, while insulating users from the complexity of the underlying systems.
Customer engagement and operational efficiency are strategic goals for every enterprise. As Alan
Trefler, CEO and founder of Pegasystems, writes with respect to attaining engagement with the
emerging generation of digitally bred customers, “There are three core principles for survival:
democratize how the organization does technology; think in layers; and use analytics.6”
The unified approach provided by Pega Platform embodies each of these principles. It democratizes
through model-based development that is accessible to both the business team as well as the technology
team. It directly builds applications as layers that match the many variations of customer and enterprise
needs. And it includes the advanced analytics needed to integrate analytics-based decision making directly
into the applications on the front lines of customer engagement.
We are Pegasystems (NASDAQ: PEGA) - dedicated to streamlining business and enhancing customer
engagement. Our Global 3000 customers rely on our dynamic, strategic applications to drive excellence in
their sales, marketing, service, and operations. We seamlessly connect our customers with their customers
across channels, in real-time.
We are dedicated to helping our clients realize exceptional customer engagement and achieve
operational excellence. Our adaptive, cloud-architected applications – built on our unified Pega®
Platform – empower people with comprehensive visual tools to extend and change applications
to meet strategic business needs. For more information visit us at www.pega.com.
© Copyright 2020 Pegasystems Inc. All rights reserved. All trademarks are the property of their respective
owners. The information contained in this press release is not a commitment, promise, or legal obligation to
deliver any material, code or functionality. The development, release and timing of any features or functionality
described remains at the sole discretion of Pegasystems. Pegasystems specifically disclaims any liability with
respect to this information.
6Trefler, Alan, Build for Change: Revolutionizing Customer Engagement through Continuous Digital Innovation, Wiley,
2014.

Pega Platform - Architecture and Technical Brief - 0

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pega Platform - Architecture and Technical Brief - 0

Uploaded by

Copyright:

Available Formats

Trademarks

This document is the property of:

1 “What is Strategy,” Michael E. Porter, Harvard Business Review, reprint 96608.

Figure 1: Pega Platform

“Model-driven” or “low-code” application development provides outstanding productivity and collaboration

Figure 2: Situational Layer Cake™

Figure 3: The three pillars

2.1 An example of a model-based strategic business application

ONBOARDING FOR RETAIL BANKING APPLICATION

Stages of Mortgage Product Case

Figure 4: Onboarding for retail banking application

Figure 5: Onboarding for retail banking models

Figure 6: Pega Platform application definition

a. An application as a group of models (a ruleset)

c. An application as layered groups of models (rulesets) with versions

d. Multiple applications from rulesets

2.3 Situation-based modeling

Figure 7a. Class Hierarchy of the Retail Onboarding cases

Figure 7b. Class definition with inheritance from two parents

2.4 Decision management and models

Figure 8: User screen showing mortgage product recommendations

Figure 9: Decision table (partial) for product recommendations

2.5 Declarative computation models

2.6 Design validation

Figure 11: Design validation guardrails

3.2 Dynamic model selection

• Models organized into classes to support specialization through inheritance

1. According to primary class inheritance

Figure 13: Dynamic model selection with rulesets

Example 2: Showing ruleset versioning

3.3 Dynamic working memory, caching, and persistence

Figure 14: The Pega Platform Engine

The working memory is dynamic in the sense that:

Figure 15: Responsive user interface – example

Figure 16b: CSR interface

4.3 Mobile offline support

• The ability to create and complete work.

Figure 18a. Easy conversation configuration and management across channels

c. Real-time and no-downtime updates deployment

4.6 Integrated social collaboration

• Customer service personnel and product managers to resolve customer issues.

Figure 19: Pega Pulse – Integrated Social Capabilities

4.8 Pega Process Extender for Salesforce Lightning

5.1 Pega® Data Integration

Figure 20: Supported integration technologies

Figure 21: Pega Live Data Integration

5.3 Pega Robotic Process Automation

Figure 22: Pega Robotic Process Automation

Figure 23: Pega Robotic Automation with Live Data integration

Figure 24: Pega Customer Decision Hub – Component Architecture

• Categorization models that assign incoming text to a predefined category:

Figure 26: Pega Cloud Services Scope

• Configures and provisions instances.

8.1 Enterprise-class security

5 OWASP top 10 security vulnerability risks: https://www.owasp.org/index.php/Top_10_2013-Top_10

8.2 Enterprise-class deployment and manageability in the cloud or on-premises

8.3 Enterprise-class reliability and scalability

• Full low-code application development providing outstanding productivity, empowering citizen

• Dynamic runtime self-assembly of applications enabling implementations unique to the specialized

9.1 The broader capabilities of Pega’s applications

You might also like