You are on page 1of 1

Prashant Jha

Data Scientist

prashantjha.jec@gmail.com 7000360764 linkedin.com/in/prashant-jha-10ab3a142 github.com/dis-is-pj

WORK EXPERIENCE TECH STACK


System Engineer - Data Languages
Python, MySQL, JavaScript, HTML
Tata Consultancy Services
01/2021 - Present, Frameworks
Tasks Scikit-learn, Pandas, Numpy, Matplotlib, Seaborn,
Led a team of four colleagues during the training period to perform and present an spaCy, NLTK, Gensim, TensorFlow2, Pytorch,
Exploratory Data Analysis task in Python on Indian Premier League ball by ball dataset. OpenCV, Flask, SQLAlchemy
Created multiple jobs in Big Data tool Talend for extraction, transformation and loading
Platforms/Tools
of data from large csv and excel files to relational database in MySQL.
Jupyter Notebook, Google Colab, Spyder, MySQL-
Workbench, PostgreSQL, Spyder, Atom, Google
Cloud Platform, GitHub, Talend, AWS, Heroku, Excel
PROJECTS
Mobile Price Range Prediction
RELEVANT COURSEWORK
AlmaBetter Verified Project
05/2021 - 06/2021, Getting Started with TensorFlow2(Coursera)
Neural Networks, ANN, CNN, Transfer Learning
Tags : Multi-Class Classification, Interpretability, Bagging, Boosting, Ensembles, Feature Engineering,
Developed a multi-class classification model using XGBoost, AdaBoost and Random
Natural Language Processing with
Forest to predict mobile price range in the market based on technical specs and
Classification and Vector Spaces (Coursera)
achieved a precision of 94% on test data and a minimum recall of 92% for each class. Sentiment Analysis, Vectorization
Carried out feature engineering themed across screen dimensions, battery life,
microprocessor strength, camera specifications, internal memory specs.
Performed correlation analysis using ANOVA and Chi-Squared tests and carried out
hyperparameter tuning using Grid Search Optimization with 5-fold cross-validation.
PUBLICATIONS
Utilized SHAP plots to identify key drivers such as RAM, battery power, mobile weight, Medium
pixel specifications and estimated an increase in the customer acquisition rate by 15%. Blogs on Data Science
2021

Company Category Classification YouTube


AlmaBetter Verified Project Vlogs on Data Science
02/2021 - 03/2021, 2021
Tags: Unsupervised Learning, Clustering, Text Processing, Gensim, Dimensionality Reduction, Word2Vec
Implemented K-means clustering on company description, meta keywords, and
homepage texts to categorize 74K companies into 12 different industries. EDUCATION
Created a single text feature by combining multiple columns representing different
areas of websites like homepage text, navbar links, headings, and meta description. B.E in Computer Science & Engg.
Employed text preprocessing techniques as text cleaning, lemmatization, stopword Jabalpur Engg. College, Jabalpur
removal, tokenization, vectorization using Gensim Word2vec and TF IDF vectorizer. 2016 - 2020,
Evaluated the optimal clusters using the Silhouette score and Elbow method to get the CGPA - 6.63
optimal number of clusters and used Word-Cloud to visualize and validate the clusters.

XII - Higher Secondary


Waste Management System
Dayal Public School, Guna, M.P.
Smart India Hackathon Winning Project 2014,
09/2020 - 10/2020,
Percentage - 89%
Tags: Computer Vision, Flask, End-to-End ML, Model Deployment, Rest-API, Cloud Computing, CNN
Build a software ecosystem to monitor and manage waste with the help of drones and
performed statistical analyses to detect the areas where more work is needed. X - Secondary
Trained an Image Classification model by using FastAI library and deployed it as a rest Saraswati Shishu Mandir, Guna, M.P.
API, which takes image address as input and classifies it into three categories. 2012,
Synthesized one-year historical waste management operations data and developed a Percentage - 86%
dashboard to monitor the performance of municipal corporations.
Used the Google Cloud platform for deploying and running the whole application, used
MySQL on GCP for live updates and used the firebase for cloud storage of the images.
Competed with more than 300 teams from engineering colleges all over India, got a
ACHIEVEMENTS
place in five finalists, and eventually won the three days long grand finale. Got in top 1% in Data Science competition
Got 36 rank out of more than 5700 participants in Machine
Learning Competition organized by MachineHack.

Coding Instructor at CueMath


Taught Coding to the school kids from U.S.

You might also like