Shree Varshan

Machine Learning, Deep Learning, Security

View on GitHub Get in Touch

CS Engineer · AI/ML · Systems · SRM University AP

scroll

About

CS student at SRM University AP (B.Tech CSE), deep in AI/ML and Generative AI. I build things most people assume need a framework: procedural sprite engines, autonomous vision agents, browser companions with zero external dependencies. Leading the technical wing of ACM SRM AP. Previously interned as a Growth & Market Researcher at Hush Networks.

Projects

cited/web-scrapper

Cited – Multi-Source Answer Aggregator

Self-hosted search engine that scrapes Stack Overflow, Hacker News, Reddit, Medium, Dev.to, and X in parallel, ranks results by source authority & recency, and feeds top passages into a RAG pipeline via Google Gemini API.

• Centralized rate-limiting with exponential backoff, retry jitter, and concurrency semaphores.
• SQLite cache for raw scraper outputs and synthesized RAG answers.
• Dynamic post-generation citation validation ensuring all inline markers match valid reference URLs.
• Docker-orchestrated backend deployment with headless Playwright for JS-heavy targets.

Python FastAPI RAG Gemini API Docker Playwright
GitHub ↗
autonomous-nav-agent

Autonomous Navigation Agent

Environment-aware autonomous agent that uses localized computer vision matrices for algorithmic pathing — built with Python & OpenCV.

Python OpenCV Computer Vision
GitHub ↗
srmap.hosting.acm.org

ACM SRM AP Website

Official site for the ACM Student Chapter at SRM AP. Built for the chapter I now lead technically.

Next.js React Tailwind
Live ↗ GitHub ↗
nakama-browser-companion

Nakama – Interactive Browser Companion

A browser extension companion that renders pixel-art sprites dynamically at runtime, with zero external image dependencies.

• Designed a custom procedural rendering engine using HTML5 Canvas math matrices to generate pixel-art sprite sheets at runtime (0B external assets).
• Engineered a game loop with a deterministic finite state machine (FSM) for 8 behavioral states and AABB collision detection.
• Deployed cross-script runtime messaging via background workers, hostname theme mapping, and localized extension storage merging.

JavaScript HTML5 Canvas Browser Extension FSM
GitHub ↗

Experience & leadership

▸ Growth & Market Researcher Intern, Hush Networks (Remote)

May 2026 – Jul 2026

• Mapped target tech ecosystems to identify growth vectors. Analyzed consumer data to guide product positioning.

▸ Technical Lead — ACM Student Chapter, SRM AP

Feb 2026 – Present · on-site

• Organized hands-on technical workshops and developer initiatives to mentor over 100+ student engineers.

▸ Co-Lead — Microsoft Student Community, SRMAP

Nov 2025 – Feb 2026 · on-site

• Led a flagship campus-wide hackathon with 500+ attendees. Directed 3 operational units to ensure a smooth event.
• Ran the full hackathon lifecycle — from planning to execution — under Microsoft Student Community.


▸ Member, Events Team

Sep 2025 – Nov 2025

• Resolved on-site technical issues quickly during live events.

Stack

core.sys

Systems & Core

  • C
  • C++
  • Python
ml.ai

AI & ML

  • PyTorch
  • Hugging Face
  • OpenCV
  • Scikit-learn
  • NumPy
dev.web

Web & Dev

  • HTML
  • CSS
  • JavaScript
  • Vercel
env.tool

Tools & Env

  • Git
  • GitHub
  • Jupyter Notebook
  • Kaggle
  • Linux

Let's build something.

Open to research collaborations, internship opportunities, and building weird, ambitious experiments with AI and systems.

shreevarshan2699@gmail.com
GitHub LinkedIn Medium X