Project Report

SAVITRIBAI PHULE PUNE UNIVERSITY,PUNE
A
PROJECT STAGE-I REPORT ON
“UNIVERSAL AI’
SUBMITTED TOWARDS THE
PARTIAL FULFILLMENT OF THE AWARD OF THE DEGREE OF
BACHELOR OF TECHNOLOGY (ECE ENGINEERING)
BY
MR. MAHESH SALPURE [UEC20M1049]

MR. YUVRAJ SHEJUL [UEC20M1051]
UNDER THE GUIDANCE OF
DR. RADHAKRISHNA NAIK

DR. D. G LOKHANDE
Department of Electronics and Computer Engineering

SANJIVANI COLLEGE OF ENGINEERING
KOPARGAON-423603
(AN AUTONOMOUS INSTITUTE)
2023-2024
SANJIVANI COLLEGE OF
ENGINEERING,KOPARGAON
INSTITUTE VISION
To Develop World Class Professionals through Quality Education.
INSTITUTE MISSION
To create Academic Excellence in the field of Engineering and Management through Educa-
tion, Training and Research to improve quality of life of people.
QUALITY POLICY
We are committed to provide quality education in the field of engineering and management
as per the norms of affiliating and accrediting authorities. We shall achieve this through
continual improvement in quality of teaching learning process, research and development.
We shall strive to achieve satisfaction of students and society.
DEPARTMENT OF ELECTRONICS AND COMPUTER ENGINEERING
VISION
To produce quality professionals in the field of Electronics and Computer Engineering with
knowledge and skill sets to meet diversifying needs of industry and society.
MISSION
• M1-To impart the technology of Electronics and Computer Engineering through effec-
tive teaching-learning process.
• M2-To establish linkages between industry and academia for overall development of
students.
• M3-To promote innovative ideas in solving multi-disciplinary engineering problems

having social relevance.
• M4-To develop technical human resources exhibiting professional and ethical attitudes.
Program Educational Objectives (PEO)
The PEOs of undergraduate program in Electronics and Computer Engineering are broadly
classified as follows:
1. PEO1: Involve in design, manufacturing, integration and testing of products, software

and systems in the field of Electronics Computer engineering and allied disciplines.
2. PEO2: Solve engineering problems having social relevance by applying knowledge

and skill sets related to Electronics and Computer engineering.
3. PEO3: Pursue higher Education/Research in the field of Engineering/Management or

become a successful entrepreneur in the related areas.
4. PEO4: Work effectively as an individual and/or a team member of multi disciplinary

assignments involving people across different cultures and national boundaries.
Program Specific Outcomes (PSO)
At the end of the program students will be able to:
• PSO1: Design, test and implement electronic systems and appliances related to signal
processing, embedded systems, industrial automation and IoT using the state-of-the-art
components and software.
• PSO2: Architect, classify and select appropriate technologies and software’s for the
implementation of computer engineering systems.
Program Outcomes (POs)
• PO1. Engineering knowledge: Apply the knowledge of mathematics, science, engi-

neering fundamentals, and an engineering specialization to the solution of complex
engineering problems.
• PO2. Problem analysis: Identify, formulate, review research literature, and analyze
complex engineering problems reaching substantiated conclusions using first princi-
ples of mathematics, natural sciences, and engineering sciences.
• PO3. Design/development of solutions: Design solutions for complex engineering

problems and design system components or processes that meet the specified needs
with appropriate consideration for the public health and safety, and the cultural, soci-
etal and environmental considerations.
• PO4. Conduct investigations of complex problems: Use research- based knowledge

and research methods including design of experiments, analysis and interpretation of
data, and synthesis of the information to provide valid conclusions.
• PO5. Modern tool usage: Create, select and apply appropriate techniques, resources,
and modern engineering and IT tools including prediction and modeling to complex
engineering activities with an understanding of the limitations.
• PO6. The engineer and society: Apply reasoning informed by the contextual knowl-
edge to assess societal, health, safety, legal and cultural issues and the consequent
responsibilities relevant to the professional engineering practice.
• PO7. Environment and sustainability: Understand the impact of the professional engi-
neering solutions in societal and environmental contexts, and demonstrate the knowl-
edge of, and need for sustainable development.
• PO8. Ethics: Apply ethical principles and commit to professional ethics and responsi-
bilities and norms of the engineering practice.
• PO9. Individual and team work: Function effectively as an individual and as a member
or leader in diverse teams, and in multidisciplinary settings.
• PO10. Communication: Communicate effectively on complex engineering activities

with the engineering community and with society at large, such as, being able to com-
prehend and write effective reports and design documentation, make effective presen-
tations, and give/ receive clear instruction.
• PO11. Project management and finance: Demonstrate knowledge and understand the
engineering and management principles and apply these to one’s own work, as a mem-
ber and leader in a team to manage projects in multidisciplinary environments.
• PO12. Life-Long Learning: Recognize the need for, and have the preparation and
ability to engage in independent and lifelong learning in the broadest context of tech-
nological change.
4
Sanjivani College of Engineering, Kopargaon-423603
(An Autonomous Institute)
DEPARTMENT OF ELECTRONICS AND COMPUTER ENGINEERING
CERTIFICATE
This is to certify that the project entitled
“UNIVERSAL AI”
Submitted by

is a bonafide work carried out by students under the supervision of

Dr. Radhakrishna Naik and Dr. D. G Lokhande and it is submitted towards the partial fulfillment of
the requirement of Bachelor of Technology (ECE Engineering).
During the Academic Year 2023-24
Dr. Radhakrishna Naik Prof. N. D. Kapale

[Internal Guide] [ Project Co-Ordinator]
Dr. B.S. Agarkar Dr. A. G. Thakur

[Head of Dept.] [ Director]
SAVITRIBAI PHULE PUNE UNIVERSITY,
PUNE
EXAMINER’S CERTIFICATE
This is to certify that the project entitled
“UNIVERSAL AI”
Submitted by

for partial fulfillment of the requirement of the award of degree of Bachelor of Technology in
Electronics Computer Engineering is examined and certified.
Signature of Internal Examiner Signature of External Examiner

Name of Sponsorer Industry with address
SPONSORER CERTIFICATE
This is to certify that the project work entitled
“‘UNIVERSAL AI”
carried out By

bonafied student of Sanjivani College of Engineering, Kopargaon 423 603 in partial

fulfillment for the award of Bachelor of Technology in Electronics Computer
Engineering of the Savitribai Phule Pune University, Pune during the academic year
2023-2024. It is certified that they has completed the project satisfactorily.
Name signature of Industry/Organization Authority

S.P.P.U, Pune SRES’s Sanjivani COE Kopargaon 423603 Departmentt of Electronics and Computer Engineering. 2023-2024
ACKNOWLEDGEMENT
“Write acknowledge to Guide, Co Guide, Project Coordinator, HOD,Director,Teachers

and supporting staff,Friends, ,Family
MR. MAHESH SALPURE

MR. YUVRAJ SHEJUL
(B.TECH. ECE)
UNIVERSAL AI I
ABSTRACT
The increasing reliance on chemical medications for common ailments has led to a surge
in healthcare concerns, prompting the need for alternative solutions. This project addresses
this challenge by developing a healthcare chatbot that not only offers users information on
traditional Indian remedies for minor ailments but also integrates essential features to en-
hance overall healthcare accessibility. The system encompasses a web application providing
details on nearby hospitals, government health schemes, and a unique health card utilizing
NFC technology to securely store and manage users’ medical history. By amalgamating
traditional wisdom with modern technology, the project aims to promote the use of natural
remedies and simplify the access to crucial healthcare information in an efficient and user-
friendly manner.
Over the past two years, significant advancements have been observed in healthcare technol-
ogy, particularly in the fields of natural language processing (NLP) and mobile health appli-
cations. Recent papers and technologies have demonstrated the effectiveness of chatbots in
providing personalized healthcare information and the integration of traditional remedies into
digital platforms. Additionally, the use of NFC technology in health cards has gained mo-
mentum, offering a secure and convenient way to store and retrieve medical records. These
developments form the foundation for our project, leveraging the latest technologies to create
a comprehensive healthcare solution that bridges the gap between traditional remedies and
modern healthcare systems.
Our contribution to this evolving landscape is a multifaceted healthcare chatbot and web ap-
plication that not only dispenses information on traditional Indian remedies but also serves
as a holistic healthcare hub. The chatbot employs advanced NLP techniques to understand
and respond to user queries, offering personalized advice on minor ailments. The web appli-
cation complements this by providing real-time information on nearby hospitals and govern-
UNIVERSAL AI II
ment health initiatives. The incorporation of an NFC-based health card ensures the secure
storage and accessibility of users’ medical histories, fostering quick and efficient response
in emergency situations. This project represents a pioneering effort to combine traditional
healthcare wisdom with contemporary technologies, promoting a more balanced and acces-
sible approach to healthcare for the modern generation.
UNIVERSAL AI III
Contents
1 INTRODUCTION 1
1.1 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 BUSINESS UNDERSTANDING 3
2.1 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Maximize: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 Minimize: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.4 Constraints: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 SOFTWARE DESIGN 6
4 Dialogue policies 8
5 Adding forms to chatbot 10
6 Implementing Fallback Policy 13
7 Deploying chatbot 17
8 Benefits to the client 19
9 Universal AI User Interface 22
REFERENCES 30
Annexure A Paper Publication 31
Annexure B Project Competition 36
Annexure C Data sheet 40
UNIVERSAL AI IV
List of Figures
3.1 Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
9.1 Paper Publication - Certificate 1 . . . . . . . . . . . . . . . . . . . . . . . . 31

9.6 Project Competition: Certificate 1 . . . . . . . . . . . . . . . . . . . . . . . 36
9.10 Data sheet1: Name of Datasheet 1 . . . . . . . . . . . . . . . . . . . . . . . 40
9.11 Data sheet2: Name of Datasheet 2 . . . . . . . . . . . . . . . . . . . . . . . 41
UNIVERSAL AI V
List of Tables
UNIVERSAL AI VI
Chapter 1
INTRODUCTION
1.1 Problem Statement
The contemporary healthcare landscape is marked by a prevailing reliance on chemical med-

ications for addressing minor ailments, leading to concerns about the potential long-term im-
plications on public health. The overuse and accessibility of pharmaceuticals raise questions
about the need for alternative solutions that could alleviate common health issues without
resorting to synthetic compounds. Additionally, the lack of centralized platforms providing
comprehensive information about nearby hospitals, government health schemes, and secure
storage of medical histories poses challenges in ensuring efficient and accessible healthcare
services. Addressing these concerns requires a reevaluation of current healthcare practices,
focusing on the development of alternative approaches that integrate traditional wisdom and
modern technology to minimize dependence on chemical medications and enhance overall
healthcare accessibility.
1.2 Scope
1. Traditional Remedies Integration:
- Focus on researching, cataloging, and integrating authentic information about tradi-

tional Indian remedies into the chatbot interface. Ensure the chatbot can effectively
provide users with advice on natural and cultural remedies for common ailments.
2. Health Information Accessibility:*
- Concentrate on developing a user-friendly web application that serves as a centralized

platform for accessing relevant healthcare information. Prioritize features such as real-
UNIVERSAL AI 1
time information on nearby hospitals, government health schemes, and an intuitive

interface for easy navigation.
3. NFC-Based Health Card Implementation:
- Prioritize the development and implementation of the NFC-based health card system,
emphasizing secure storage and easy retrieval of users’ medical histories. Ensure the
health card is portable, password-protected, and includes key information for emer-
gency situations. Focus on the seamless integration of this feature into the overall
project architecture.
4. Educational Content Development:
- Develop informative content within the chatbot and web application to educate users
about traditional Indian remedies. - Emphasize the benefits, limitations, and appropri-
ate use of traditional remedies, promoting informed decision-making and a balanced
approach to healthcare.
1.3 Objectives
• 1. Promote Holistic Healthcare:
- Increase awareness and adoption of traditional Indian remedies for common ailments
through an intelligent chatbot interface. - Encourage a balanced healthcare approach,
reducing dependence on chemical medications for minor health issues.
• 2. Enhance Accessibility:
- Develop a user-friendly web application offering information on nearby hospitals and

government health schemes. - Facilitate quick access to relevant healthcare resources
for informed decision-making.
• 3. Secure Health Information:
- Implement an NFC-based health card system for secure storage and easy retrieval
of users’ medical histories. - Provide a portable, password-protected health card with
essential information for emergency response.
UNIVERSAL AI 2
Chapter 2
BUSINESS UNDERSTANDING
2.1 Problem Statement
The contemporary healthcare landscape is marked by a prevailing reliance on chemical med-

ications for addressing minor ailments, leading to concerns about the potential long-term im-
plications on public health. The overuse and accessibility of pharmaceuticals raise questions
about the need for alternative solutions that could alleviate common health issues without
resorting to synthetic compounds. Additionally, the lack of centralized platforms providing
comprehensive information about nearby hospitals, government health schemes, and secure
storage of medical histories poses challenges in ensuring efficient and accessible healthcare
services. Addressing these concerns requires a reevaluation of current healthcare practices,
focusing on the development of alternative approaches that integrate traditional wisdom and
modern technology to minimize dependence on chemical medications and enhance overall
healthcare accessibility.
2.2 Maximize:
1. User Engagement:
- KPI: Increase in the number of user interactions with the chatbot and web application. -
Strategy: Implement engaging and informative content within the chatbot and web appli-
cation to encourage frequent usage. Track user sessions, queries, and feedback to assess
engagement levels.
2. Adoption of Traditional Remedies:
- KPI: Percentage increase in the utilization of traditional Indian remedies suggested by the
chatbot. - Strategy: Regularly update and expand the traditional remedy database. Encourage
UNIVERSAL AI 3
user testimonials and success stories related to the effectiveness of traditional remedies.
3. Accessibility to Healthcare Resources:
- KPI: Reduction in the time taken for users to access information about nearby hospitals
and government health schemes. - Strategy: Optimize geolocation services for quick and ac-
curate information retrieval. Ensure the web application’s user interface facilitates seamless
navigation.
4. Health Card Usage:
- KPI: Growth in the number of users utilizing the NFC-based health card system. - Strategy:
Emphasize the convenience and security of the health card. Implement user-friendly features
and incentives to encourage users to maintain and utilize their digital health cards.
2.3 Minimize:
1. Reliance on Chemical Medications:

-KPI: Decrease in the frequency of user queries related to chemical medications for minor ail-
ments. - Strategy: Educate users about the potential side effects of chemical medications for
common health concerns. Highlight the benefits and safety of traditional remedies through
the chatbot and educational content.
2. Health Information Retrieval Time:
- KPI: Reduction in the time taken for emergency responders to access critical health infor-
mation through the health card. - Strategy: Optimize the health card system for rapid data
retrieval. Implement secure yet efficient authentication processes for emergency situations.
3. Physical Documentation Reduction:
- KPI: Decrease in the reliance on physical documentation for medical histories. - Strategy:
Promote the use of the digital health card as a secure and portable alternative to traditional
paper-based medical records. Emphasize the environmental benefits and convenience of dig-
ital documentation.
UNIVERSAL AI 4
2.4 Constraints:
1. Technological Limitations:
- The effectiveness of the solution is contingent on the availability and reliability of internet
connectivity for users interacting with the chatbot and accessing the web application.
2. Privacy and Security Concerns:
- Adherence to stringent data protection regulations is imperative to ensure the confidentiality
and security of users’ health information stored in the digital health card.
3. User Resistance to Traditional Remedies:
- Some users may be resistant to adopting traditional Indian remedies, requiring a nuanced
approach to education and information dissemination.
4. Device Compatibility:
- The success of the health card system relies on users having NFC-enabled devices, poten-
tially limiting accessibility for individuals with older or incompatible devices.
UNIVERSAL AI 5
Chapter 3
SOFTWARE DESIGN
Block diagram and its explanation:

7.1 User:
Implements user-related actions, including Register() method for user registration and incor-
porating Login() method for user authentication.
7.2 Dashboard :
Central hub with various features, containing HealthcareSchemes() for navigating to health-
care schemes, providing BookAppointment() for doctor appointment booking, including
TapToChat() for chatbot interactions, and featuring TapToAsk() for asking questions using
voice commands.
7.3 HealthcareSchemes :
Manages the display of disease categories and government schemes, including viewDisease-
Categories() method for showing available disease categories and incorporating viewGov-
ernmentSchemes() method for displaying government healthcare schemes.
7.4 DiseaseCategory :
Represents specific disease categories and includes viewSchemes() method for displaying
government schemes related to the selected disease category.
7.5 BookAppointment :
Manages doctor appointment booking with methods for choosing a doctor, selecting date and
time, and confirming the appointment.
7.6 TapToChat :
Enables chat interactions with the chatbot, featuring chatWithChatbot() and getHomeReme-
dies() methods.
UNIVERSAL AI 6
7.7 TapToAsk :
Enables voice-based question interaction with askWithVoice() method and provides home
remedies information through getHomeRemedies() method.
7.8 PMJAY :
Represents the Pradhan Mantri Jan Arogya Yojana (PMJAY) government healthcare scheme,
featuring getDetails() method for retrieving scheme details.
7.9 AyushmanBharat :
Represents the Ayushman Bharat government healthcare scheme, featuring getDetails() method
for retrieving scheme details.
7.10 HomeRemedies :
Provides home remedies information, including displayRemedies() method for showing reme-
dies in both chat and voice interactions.
Figure 3.1: Block diagram
UNIVERSAL AI 7
Chapter 4
Dialogue policies
In a scenario where the chatbot may not actively collect and utilize user feedback for self-
improvement, alternative dialogue policies are implemented to enhance the user experience.
The chatbot relies on predefined rules, intent recognition, and contextual understanding to
provide accurate and relevant information. Below are key dialogue policies designed to op-
timize user interactions without incorporating active user feedback for improvement:
1. Intent Recognition and Contextual Understanding:

- Inference: The chatbot heavily relies on robust intent recognition mechanisms and contex-
tual understanding. By analyzing the user’s queries, the chatbot infers the user’s intent and
maintains context throughout the conversation. For example, if a user asks about remedies for
a specific ailment, the chatbot infers the intent and tailors subsequent responses accordingly.
2. Rule-Based Responses:
- Inference: The chatbot follows predefined rules to generate responses. Common user
queries related to traditional remedies and health information have predetermined responses,
ensuring accuracy and reliability. The inference is based on the recognition of key phrases
and patterns within user input.
3. Fallback Mechanism:
- Inference: When faced with ambiguous queries or user inputs outside its knowledge scope,
the chatbot activates a fallback mechanism. This mechanism infers the need for clarification
UNIVERSAL AI 8
and prompts the user with predefined clarifying questions to gather additional information
and refine its responses.
4. Dynamic Content Presentation:

- Inference: The chatbot dynamically adjusts the presentation of information based on the
user’s interaction patterns. For instance, if a user consistently asks for detailed explanations,
the chatbot infers a preference for more comprehensive information, leading to the delivery
of more in-depth responses.
5. Emergency Response Protocol:

- Inference: To handle potential emergency situations, the chatbot incorporates predefined
protocols. If the chatbot detects keywords or phrases indicative of a medical emergency,
it infers the urgency and immediately provides standard emergency response information,
including contact details for local emergency services.
6. User Engagement Monitoring:

- Inference: The chatbot monitors user engagement through session durations and interac-
tion frequency. If there is a decline in user engagement, the chatbot infers the need for re-
engagement strategies. This may involve introducing interactive elements, such as quizzes
or suggestions, to maintain user interest.
7. Progressive Disclosure of Information:

- Inference: The chatbot uses progressive disclosure to manage the flow of information. If
a user’s queries indicate a gradual understanding or interest in a specific topic, the chatbot
infers the need to progressively disclose more detailed information, ensuring that responses
align with the user’s evolving comprehension.
UNIVERSAL AI 9
Chapter 5
Adding forms to chatbot
Health Profile Form:

Universal-AI: Greetings! To provide you with the most personalized recommendations for
a headache remedy, I’d like to gather some information about your health. Your privacy is
important, and all information will be kept confidential.
1. Universal-AI: Do you currently have any existing medical conditions or chronic illnesses?
Please reply with ”Yes” or ”No.”
- If the user responds with ”Yes,” Universal-AI follows up with specific questions about the
medical conditions. For example, ”Could you please provide details about your medical
conditions, including diagnoses and any prescribed medications?”
- If the user responds with ”No,” Universal-AI proceeds to the next question.
2. Universal-AI: Are you currently taking any medications? Your response will help me
better understand your health context. Please reply with ”Yes” or ”No.”
- If the user responds with ”Yes,” Universal-AI may ask for the names of medications, their
dosage, and the frequency of use. For example, ”Could you please provide the names of the
medications you are currently taking, along with the dosage and frequency?”
3. Universal-AI: Do you have any known allergies? Understanding your allergies will help
me tailor recommendations that suit your health profile. Please reply with ”Yes” or ”No.”
UNIVERSAL AI 10
- If the user responds with ”Yes,” Universal-AI provides a list of common allergens and asks
the user to select relevant ones. For example, ”Please select any allergies from the following
list: [List of common allergens].”
4. Universal-AI: Are there any specific preferences or restrictions you have regarding reme-
dies? For example, do you prefer herbal solutions or have any dietary restrictions? Please
provide details, or reply with ”None” if there are no specific preferences.
- Universal-AI allows the user to provide additional information or preferences. For instance,
the user might say, ”I prefer herbal remedies” or ”I am vegetarian, so no animal products,
please.”
5. Universal-AI: Finally, is there anything else you’d like to share about your health that
could be relevant for personalized recommendations?
- Universal-AI provides an open-ended space for the user to share any additional information
that could impact the recommendations.
Importance:
1. Personalization and Tailoring:

- The form facilitates personalization by collecting specific user information, allowing Universal-
AI to tailor recommendations for traditional remedies based on individual health profiles.
2. Data-Driven Decision Making:
- Structured data empowers Universal-AI to make data-driven decisions, enhancing its un-
derstanding of users’ health profiles and enabling more precise recommendations.
3. User Empowerment:
- Users actively participate in their healthcare journey by providing detailed health infor-
mation, fostering a sense of control over the recommendations they receive for managing
headaches.
4. Holistic Healthcare Approach:
UNIVERSAL AI 11
- The form supports a holistic approach by considering broader health aspects, such as exist-
ing conditions, medications, allergies, and preferences, not just the specific ailment.
5. Improved Accuracy and Relevance:
- The form enhances the accuracy and relevance of recommendations, allowing Universal-
AI to provide targeted advice based on individual health profiles, contributing to a more
satisfactory user experience.
UNIVERSAL AI 12
Chapter 6
Implementing Fallback Policy
In the implementation of the fallback policy for our healthcare chatbot, it is crucial to address
the nature of the data that forms the foundation of our system. The data at hand is diverse,
encompassing a wide range of user queries related to health concerns, traditional remedies,
and general medical information. The nature of this dataset can be characterized as moder-
ately imbalanced, with a larger proportion of queries related to general health information
and a smaller but significant proportion concerning traditional remedies. This imbalance ne-
cessitates a robust fallback policy to handle user inputs that may not align with the primary
objectives of the chatbot.
To effectively manage this data, a fallback policy is implemented to gracefully handle queries
that fall outside the scope of the chatbot’s primary functions. This includes instances where
users ask questions unrelated to healthcare, input ambiguous queries, or inquire about topics
beyond the system’s expertise. The fallback policy aims to provide users with a satisfactory
response or guide them to rephrase their query to obtain more accurate information.
Criteria for Model Selection:
1. PyTorch Neural Network:

- Nature of Model: PyTorch Neural Network is chosen for its adaptability and ability to
handle various types of data. This model is suitable for our chatbot as it can be trained on the
diverse dataset, adjusting its weights to accommodate the imbalances in the data.
- Strengths: PyTorch offers a dynamic computational graph, making it well-suited for NLP
tasks. Its flexibility allows us to customize the neural network architecture, enabling effective
learning from the imbalanced data.
UNIVERSAL AI 13
2. BERT Model:
- Nature of Model: BERT (Bidirectional Encoder Representations from Transformers) is
selected due to its proficiency in understanding contextual nuances in language. It excels in
capturing intricate relationships within the dataset, which is essential for providing accurate
responses to user queries.
- Strengths: BERT’s bidirectional attention mechanism ensures a deep understanding of con-
text, making it well-suited for our chatbot’s diverse dataset. Its pre-trained embeddings cap-
ture semantic meaning, enhancing the model’s ability to handle imbalanced data effectively.
3. GPT-3 Model:
- Nature of Model: GPT-3 (Generative Pre-trained Transformer 3) is chosen for its powerful
natural language generation capabilities. GPT-3 can comprehend and generate human-like
responses, which is beneficial for crafting nuanced fallback responses to varied user inputs.
- Strengths: GPT-3’s large-scale pre-training on diverse datasets makes it proficient in under-
standing and generating contextually relevant language. Its versatility allows the chatbot to
handle a broad spectrum of user queries with finesse.
4. Rasa NLU Model:
- Nature of Model: Rasa NLU (Natural Language Understanding) is included for its ability
to interpret user intents and entities, providing a structured understanding of user input. It
contributes to the chatbot’s overall comprehension of user queries.
- Strengths: Rasa NLU excels in handling imbalanced data by allowing the definition of
custom intents and entities. Its open-source nature facilitates seamless integration with our
existing chatbot infrastructure.
The selection of these models is driven by the need for a diverse yet specialized approach
to handle the imbalances in the healthcare chatbot’s dataset. Each model contributes unique
strengths to ensure a comprehensive and adaptive fallback policy, enhancing the chatbot’s
robustness in addressing a wide array of user queries.
UNIVERSAL AI 14
Example of the Rasa Intent
NLU:
- Intent: greet
– Hello
– Hi
– Hey
- Intent: headache
– I have a headache
– How to treat a headache
– Remedies for headache
- Intent: flu
– Symptoms of flu
– Flu remedies
– I think I have the flu
- Intent: runny nose
– How to stop a runny nose

– Runny nose remedies
– I can’t stop sneezing
- Intent: asthma
– Asthma symptoms
– Managing asthma attacks
– Inhaler usage tips
- Intent: cough
– Persistent cough
UNIVERSAL AI 15
– Home remedies for cough

– When to see a doctor for a cough
- Intent: sore throat
– Sore throat remedies

– What to do for a sore throat
– Throat pain relief
- Intent: pimples
– How to get rid of pimples

– Acne treatment
– Pimple prevention tips
- Intent: anxiety
– Dealing with anxiety

– Coping strategies for anxiety
– Anxiety symptoms
UNIVERSAL AI 16
Chapter 7
Deploying chatbot
Introduction:
In the deployment phase of our healthcare chatbot project, we leverage versatile platforms to
ensure seamless accessibility and efficient integration. Three distinct deployment methods
are explored, utilizing Railway, Pinecone, and port forwarding with ngrok to host and make
our chatbot available for users.
1. Railway Deployment:
Railway serves as a robust platform for hosting web applications, simplifying the deployment
process. The integration of our healthcare chatbot involves:
- Configuration: Utilize Railway’s intuitive configuration tools to set up the necessary envi-
ronment and dependencies for the chatbot application.
- Deployment: Leverage Railway’s deployment features to host the chatbot application, mak-
ing it accessible to users via the internet.
- Scaling: Take advantage of Railway’s scalability features, ensuring optimal performance as
user demand fluctuates.
2. Pinecone Deployment:
Pinecone specializes in deploying and managing machine learning models at scale, making
it an ideal choice for deploying our chatbot. The deployment process encompasses:
- Model Integration: Integrate our chosen models, such as the PyTorch Neural Network,
BERT Model, or others, into Pinecone for efficient deployment.
- Indexing: Leverage Pinecone’s indexing capabilities to handle diverse data inputs, enhanc-
ing query response times.
- Scaling: Benefit from Pinecone’s seamless scaling capabilities, ensuring the chatbot can
efficiently handle varying loads of user queries.
UNIVERSAL AI 17
3. Port Forwarding with ngrok:

For local development and testing purposes, we use ngrok to create secure tunnels, making
our chatbot accessible over the internet. Although not recommended for production use due
to security considerations, ngrok facilitates quick testing during the development phase:
- Local Server Setup: Run the chatbot application on a local server for testing and develop-
ment purposes.
- ngrok Configuration: Utilize ngrok to create secure tunnels, exposing the local server and
making the chatbot accessible in real-time.
- Real-time Testing: Share the ngrok-generated URL for collaborative and real-time testing
of the deployed chatbot.
By adopting these platforms for deployment, we ensure a versatile and effective approach that
aligns with our project’s goals. Railway simplifies web application hosting, Pinecone excels
in deploying machine learning models, and ngrok provides a convenient solution for local
development and testing. The chosen deployment methods cater to the project’s scalability
needs, ease of use, and security considerations, ensuring an optimal user experience.
UNIVERSAL AI 18
Chapter 8
Benefits to the client
Benefits to the Client: Enhancing Healthcare through Innovative Solutions

The intelligent healthcare chatbot, with its associated services, enhances healthcare by im-
proving accessibility, user experience, and fostering a holistic approach to well-being.
1. Empowering Users with Ayurvedic Wisdom:
The healthcare chatbot provides users with access to a wealth of ayurvedic knowledge and
remedies for common ailments. By leveraging traditional Indian medicine, users can explore
holistic alternatives to conventional treatments. This not only fosters a cultural connection to
wellness but also encourages a shift towards natural and sustainable healthcare practices.
2. Reducing Dependency on Chemical Medications:
One of the primary objectives of the chatbot is to minimize the unnecessary use of chemical
medications for minor ailments. By promoting traditional remedies rooted in Ayurveda,
users are empowered to address common health issues with natural solutions. This reduction
in the reliance on pharmaceuticals aligns with the global trend towards holistic well-being
and contributes to the client’s commitment to sustainable healthcare practices.
3. Facilitating Informed Decision-Making:
The integration of a user-friendly web application complements the chatbot by providing
valuable information on nearby hospitals and government health schemes. This empowers
users to make informed decisions about their healthcare journey. The ability to access rele-
vant healthcare resources, coupled with the chatbot’s personalized advice, ensures that users
can navigate the complexities of the healthcare landscape more effectively.
4. Efficient and Secure Medical Documentation:
UNIVERSAL AI 19
The inclusion of an NFC-based health card revolutionizes the way users manage their medi-
cal records. By offering a password-protected digital repository of medical history and docu-
ments, users can seamlessly carry essential health information. In emergency situations, this
facilitates swift and accurate medical interventions, as healthcare providers gain immediate
access to critical details such as blood group, allergies, and past treatments.
5. Enhancing Emergency Response:
The health card, with its visible basic information and immediate accessibility, serves as a
crucial tool in emergency situations. In the event of accidents or sudden health crises, indi-
viduals can be quickly identified, and emergency contacts can be notified. This accelerates
the response time of healthcare professionals, contributing to more effective and potentially
life-saving interventions.
6. Promoting Holistic Healthcare Practices:
The chatbot’s focus on ayurvedic remedies, coupled with information on nearby hospitals
and government health schemes, promotes a holistic approach to healthcare. Users are en-
couraged to consider not only immediate symptom relief but also long-term well-being. This
aligns with the client’s commitment to fostering a healthcare ecosystem that values preven-
tive measures and overall health maintenance.
7. User-Centric and Accessible Services:
The combination of a web application, mobile application, and chatbot ensures that health-
care services are accessible to a diverse user base. Whether users prefer the convenience of
a mobile app, the comprehensive information on the web app, or the interactive chatbot in-
terface, the client’s healthcare ecosystem caters to varied preferences, enhancing overall user
satisfaction.
8. Technology-driven Innovation:
The incorporation of advanced technologies such as NLP, machine learning, and NFC tech-
nology positions the client as an innovator in the healthcare sector. By embracing technology,
the client not only addresses current healthcare challenges but also demonstrates a commit-
ment to staying at the forefront of industry advancements.
9. Personalized Health Advice:
The chatbot, powered by machine learning algorithms, delivers personalized health advice
based on user inputs and historical interactions. This personalized approach ensures that
UNIVERSAL AI 20
users receive recommendations tailored to their specific health conditions, symptoms, and
preferences, fostering a more engaging and effective healthcare experience.
10. User Education and Awareness:
The chatbot serves as an educational tool by providing information on traditional remedies.
Users gain awareness of alternative healthcare practices rooted in Ayurveda, contributing to
their health literacy. This focus on education aligns with the client’s commitment to empow-
ering users to make informed decisions about their well-being.
11. Cost-effective Healthcare:
By promoting natural remedies and reducing the unnecessary use of chemical medications,
the healthcare chatbot contributes to cost-effective healthcare solutions. Users can manage
minor ailments without the need for expensive pharmaceuticals, aligning with the broader
industry trend towards sustainable and affordable healthcare.
12. Enhanced Healthcare Accessibility:
The web application’s feature providing information on nearby hospitals enhances health-
care accessibility. Users can quickly locate healthcare facilities in their vicinity, facilitating
prompt medical attention when needed. This aspect is particularly valuable in emergency
situations where timely access to healthcare resources is critical.
13. Seamless Integration with Traditional Medicine:
The emphasis on ayurvedic remedies bridges the gap between modern healthcare practices
and traditional medicine. The chatbot acts as a catalyst for integrating the best of both worlds,
fostering a collaborative healthcare approach that values the strengths of traditional healing
methods alongside contemporary medical practices.
14. Improved Health Outcomes:
The holistic healthcare approach advocated by the chatbot and associated services contributes
to improved health outcomes. Users are encouraged to address health issues at their roots,
promoting overall well-being rather than merely symptomatic relief. This approach aligns
with the growing preference for preventive healthcare strategies.
UNIVERSAL AI 21
Chapter 9
Universal AI User Interface
Login Page of Mobile App
UNIVERSAL AI 22
Login Page of Web App
UNIVERSAL AI 23
Login Page Code
UNIVERSAL AI 24
Dashboard
UNIVERSAL AI 25
ChatBot
UNIVERSAL AI 26
ChatBot
UNIVERSAL AI 27
ChatBot Code
UNIVERSAL AI 28
Goverment Schemes Section
UNIVERSAL AI 29
References
[1] Wedaduru -An Intelligent Ayurvedic Disease Screening and Remedy Analysis Solution
Publisher: IEEE.
[2] Ayurvedic Diagnosis using Machine Learning Techniques to examine the diseases by
extracting the data stored in AyurDataMart Publisher: IEEE
[3] Machine Learning Applications in Healthcare Publisher: IEEE
[4] Machine Learning Applications in Healthcare
[5] Chatbot For Healthcare Using Machine Learning
UNIVERSAL AI 30
Annexure A
Paper Publication
60.jpg
Figure 9.1: Paper Publication - Certificate 1
UNIVERSAL AI 31
61.jpg
UNIVERSAL AI 32
62.jpg
UNIVERSAL AI 33
63.jpg
UNIVERSAL AI 34
64.jpg
UNIVERSAL AI 35
Annexure B
Project Competition
65.jpg
Figure 9.6: Project Competition: Certificate 1
UNIVERSAL AI 36
66.jpg
UNIVERSAL AI 37
67.jpg
UNIVERSAL AI 38
68.jpg
UNIVERSAL AI 39
Annexure C
Data sheet1
65.jpg
Figure 9.10: Data sheet1: Name of Datasheet 1
UNIVERSAL AI 40
66.jpg
Figure 9.11: Data sheet2: Name of Datasheet 2
UNIVERSAL AI 41

Project Report

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

Project Report

Uploaded by

Copyright:

Available Formats

SAVITRIBAI PHULE PUNE UNIVERSITY,PUNE

MR. MAHESH SALPURE [UEC20M1049]

UNDER THE GUIDANCE OF

DR. RADHAKRISHNA NAIK

Department of Electronics and Computer Engineering

To Develop World Class Professionals through Quality Education.

• M3-To promote innovative ideas in solving multi-disciplinary engineering problems

Program Educational Objectives (PEO)

1. PEO1: Involve in design, manufacturing, integration and testing of products, software

2. PEO2: Solve engineering problems having social relevance by applying knowledge

3. PEO3: Pursue higher Education/Research in the field of Engineering/Management or

4. PEO4: Work effectively as an individual and/or a team member of multi disciplinary

Program Outcomes (POs)

• PO1. Engineering knowledge: Apply the knowledge of mathematics, science, engi-

• PO3. Design/development of solutions: Design solutions for complex engineering

• PO4. Conduct investigations of complex problems: Use research- based knowledge

• PO10. Communication: Communicate effectively on complex engineering activities

MR. MAHESH SALPURE [UEC20M1049]

is a bonafide work carried out by students under the supervision of

Dr. Radhakrishna Naik Prof. N. D. Kapale

Dr. B.S. Agarkar Dr. A. G. Thakur

MR. MAHESH SALPURE [UEC20M1049]

Signature of Internal Examiner Signature of External Examiner

MR. MAHESH SALPURE [UEC20M1049]

bonafied student of Sanjivani College of Engineering, Kopargaon 423 603 in partial

Name signature of Industry/Organization Authority

“Write acknowledge to Guide, Co Guide, Project Coordinator, HOD,Director,Teachers

MR. MAHESH SALPURE

3.1 Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

9.1 Paper Publication - Certificate 1 . . . . . . . . . . . . . . . . . . . . . . . . 31

1.1 Problem Statement

The contemporary healthcare landscape is marked by a prevailing reliance on chemical med-

1. Traditional Remedies Integration:

- Focus on researching, cataloging, and integrating authentic information about tradi-

2. Health Information Accessibility:*

- Concentrate on developing a user-friendly web application that serves as a centralized

time information on nearby hospitals, government health schemes, and an intuitive

3. NFC-Based Health Card Implementation:

4. Educational Content Development:

• 1. Promote Holistic Healthcare:

- Develop a user-friendly web application offering information on nearby hospitals and

• 3. Secure Health Information:

2.1 Problem Statement

The contemporary healthcare landscape is marked by a prevailing reliance on chemical med-

1. Reliance on Chemical Medications:

Block diagram and its explanation:

Figure 3.1: Block diagram

1. Intent Recognition and Contextual Understanding:

4. Dynamic Content Presentation:

5. Emergency Response Protocol:

6. User Engagement Monitoring:

7. Progressive Disclosure of Information:

Adding forms to chatbot

Health Profile Form:

1. Personalization and Tailoring:

Implementing Fallback Policy

Criteria for Model Selection:

1. PyTorch Neural Network:

Example of the Rasa Intent

- Intent: runny nose

– How to stop a runny nose

– Home remedies for cough

- Intent: sore throat