Rajat's Projects

Personal Projects

Inflow VIP

Language(s)/Stack: Social Media, Marketing, Business

Inflow VIP is a social media platform that connects people with similar interests and hobbies. It allows users to share their passions, discover new interests, and engage with like-minded individuals. The platform is designed to foster a sense of community and encourage meaningful interactions among its users.

April 2025

RajatAI

Language(s)/Stack: Python (Flask), JavaScript, Heroku

Made a personal website with an AI resume bot based on ChatGPT 3.5 - Turbo. Integrated CI/CD pipeline with Heroku. I was using AWS at first, but Heroku was much faster for CI/CD. If you are reading this, you are currently using this project. Enjoy your time here :)

September 2023

Derivative Securities Portfolio Optimization

Language(s)/Stack: Matlab

Designed a portfolio of fixed-income instruments (APAC Bonds) immune to Delta and Gamma risks. Optimized the portfolio with an iterative algorithm by reimmunizing its composition using duration and convexity calculations. Mitigated macroeconomic risks by calculating interest rate term structures, extracting yield curves using the Nelson-Siegel model, and simulating hypothetical recessions.

May 2023

Machine Learning Trend Analysis Using Big Data Techniques

Language(s)/Stack: Bash, Python, Hadoop, Spark, and MySQL

Employed big data programming tools to analyze historical versions of Wikipedia’s machine learning page. Implemented MapReduce jobs to dissect co-occurrences and n-grams in the articles, further enhancing the depth of analysis. Transformed raw data into actionable insights, presenting the evolution of machine learning concepts, techniques, and milestones.

March 2023

Artificial Intelligence Models Marketplace

Language(s)/Stack: Java, JavaScript, MySQL, Spring Boot, and Maven

Created a full stack Marketplace selling Artificial Intelligence models. Customers can create an account, log in/out, view available AI models, add them to their shopping cart, checkout from shopping cart, and view their order history. Admin User can add/update/delete/hide AI models, view orders, change order status.

March 2023

Pacman AI

Language(s)/Stack: Python

Engineered an Artificial Intelligence to play Pacman. Based on UC Berkeley’s AI course. Used probabilistic inference in a hidden Markov model to track ghosts' movements. Implemented machine learning classification algorithms such as Naive Bayes and Reinforcement Learning.

December 2022

Movie Recommendation Website

Language(s)/Stack: Python (Django), MySQL

Developed a full-stack movie recommending website. Utilized a cosine similarity algorithm to match each user with their “movie soulmate” by calculating which other user has the most similar movie interests. Users can like, dislike, and comment on each movie. Admin users can add movies, edit movies, and delete offensive comments.

May 2022