Varun Sadasivam

COMPUTER SCIENCE @ UNIVERSITY OF MICHIGAN

Dive into my world of programming, chess, and outer space as I continue to explore the tech world!

Contact Me

Recent Projects

Chessir - AI Chess Trainer

Chessir - AI Chess Trainer

NASA S.U.I.T.S

NASA S.U.I.T.S

Helix - Agentic AI Assistant

Helix - Agentic AI Assistant

Experience

NASA S.U.I.T.S - CLAWS UM 🚀 👨🏽‍🚀 🪐

https://claws.engin.umich.edu
ReactJS   ExpressJS   Python Flask   MongoDB   WebSocket   AWS   HoloLens

I led a team of 20+ engineers to develop a full-stack application with augmented reality integration for NASA astronauts, using the Microsoft HoloLens. The project involved building a scalable backend with Python Flask, WebSocket protocols, MongoDB, Redis, and AWS, alongside a ReactJS frontend designed for reliable performance on unstable networks. We optimized our development pipeline with CI/CD GitHub Actions and Kubernetes and implemented key features like route optimization algorithms and AR synchronization. Comprehensive testing was conducted using Jest, Cypress, and Pytest to maintain high code quality. We worked with NASA engineers and presented our work at the Johnson Space Center in Houston, Texas!

New York Life Insurance - Software Engineering Intern 🗽 🍎 💻

ReactJS   Java Spring Boot

This internship allowed me to get both technical and business experience on a software product built on technologies like ReactJS and Java Spring Boot. I worked on a team which developed NYL's internal consulting software, GuideMe, which is used by over 13,000 agents and managers to help in the sales process.

Over the course of the program, I developed an accessibility feature which vastly improved visuals for those with colorblindness and other visual impairments, improved screen-reader support, and implemented video summarizing solutions. I also got experience with automated testing by using Selenium which vastly helped in the process. This experience was a great blend of technical and business skills and I learned a lot about how large softwares are developed and maintained. This was also my first time in NYC, and I really fell in love with the city!

HBK Engineering - Software Engineering Intern 🧑🏽‍💻 🛠️ 🚧

C++   Python   JavaScript   Qt

In this internship I got a chance to work on a couple of different projects. The first one was a multi-threaded desktop application written in C++ and Qt which streamlined financial data parsing, making the process much faster and more efficient. I also got to work on a JavaScript app that seamlessly worked with a vast collection of electric pole diagrams and spreadsheets, dramatically speeding up data conversion. Lastly, I designed a central dashboard using PowerBI and Python that brought all key metrics and KPIs together in one place, enhancing team communication and collaboration.

This role was a great blend of problem-solving and innovation, allowing me to dive into diverse technologies and make a real impact on how data is managed and visualized.

Projects

Chessir - AI Chess Trainer 🤖 🦾 ♟️

https://www.chessir.com
ReactJS   Python Quart   Redis   Web Workers   Celery   MongoDB   OAuth

A web app that enhances chess skills by generating personalized puzzles based on players' Chess.com games. The application utilizes a custom algorithm interfacing with the AI engine Stockfish to identify missed winning moves and fluctuations in game data, producing tailored puzzles for users. By integrating with the Chess.com API, the site allows authenticated users to select and analyze their games, receiving puzzles stored in a database for targeted practice.

I have partnered with chess academies to help improve students' strategic thinking and performance through personalized training. It was a tough project which taught me about RESTful APIs, OAuth, and data analysis and it was amazing seeing students improve using it!

Helix - Agentic AI Assistant 🧬 🤖 🧑🏽‍💻️

Langgraph   ReactJS   Python Flask   WebSocket   SQL

This is a conversational agentic AI platform built to be an HR assistant for generating recruiting outreach steps. The backend uses Langgraph and Langchain to build the agentic infrastructure which is served over Flask, communicated via WebSocket, and stored with SQL. This was my first project learning about agentic AI and I enjoyed creating the nodes and edges to form a graph using Langgraph. The AI is able to make decisions using conditional edges which direct which nodes it lands on.

headshot image

More about me

I grew up in Chicago, went to high school at Walter Payton and now study at the University of Michigan. I first got into programming when I tried to develop cheats for a game called Minecraft when I was younger and I have carried that love for tinkering with me. Outside of programming, I like to tutor kids in chess, learn about philosophy and religion, and play Spikeball on the beach!

Contact Me

Socials:

LinkedinGithub
contact illustration

Varun Sadasivam

© All rights are reserved | 2025
Github