snake.yml workflow below runs once — see setup note at the bottom
name: Mitul Bhatia
role: AI / Full-Stack Engineer — Agentic Systems & LLM Infra
education: B.Tech Artificial Intelligence, Newton School of Technology (Rishihood University)
cgpa: 9.80 / 10.0
based_in: Sonipat, Haryana 🇮🇳
foundation: 1400+ LeetCode · 900+ Codeforces
mindset: product-first engineering, ship real systems not toy apps🚀 Currently building → raindeer.social — an AI-first social media OS for creators, solo founders & small teams, powered by a multi-agent LangGraph pipeline with persistent brand memory.
🧠 Deep in → multi-agent orchestration, RAG pipelines, guardrails & AI security, and full-stack shipping (FastAPI + Next.js + PostgreSQL).
🤝 Open to collaborate on → agentic systems, backend architecture, and real product ideas — the kind that ship, not the kind that stay in a Figma file.
💬 Ask me about → LangGraph pipelines, RAG architecture, system design, or building "the next idea for Silicon Valley."
⚡ Fun fact → I'm always equipped with a 'WE BALL' kind of attitude — ship first, perfect later.
I contribute to open-source under real collaboration standards — resolving issues and getting PRs reviewed and merged (Hacktoberfest), and I keep my own builds public so others can fork, learn from, or extend them. agrovers_wow and rag-web-search-agent are good entry points if you want to poke around RAG/agent code — issues and PRs are genuinely welcome.
|
Multi-Agent AI Security & Guardrail Engine Autonomous 7-agent pipeline that detects, exploits, patches, and validates software vulnerabilities with zero human intervention. Real-time AST policy evaluation and prompt-injection guardrails running inside Docker — sub-24ms security triage latency with 99.8% injection recall.
|
Real-Time Roommate & Housing Matching System Full-stack housing/roommate matching app with multi-factor preference scoring and low-latency state delivery. Horizontal Redis Pub/Sub scaling across WebSocket workers — sub-45ms sync latency, 94.2% match precision.
|
|
AI-Powered Soil & Crop Intelligence Assistant AI assistant that converts raw soil inputs into crop/fertilizer recommendations for low-connectivity, low-literacy users. Multi-stage validation on top of FAISS-based RAG — improved recommendation reliability by 35%.
|
Always shipping something new Explore the rest of the build log — RAG search agents, race-strategy data science, and everything else in progress.
|
Languages
AI / Agentic Systems
Backend & Infra
Frontend
Data & ML
Tools, Design & Deploy
| 🧮 LeetCode | 1400+ rating |
| ⚔️ Codeforces | 900+ rating |
| 🎃 Hacktoberfest | Contributor — resolved issues & merged PRs under real collab standards |
| 🚩 Hackathon | Mumbai Hacks — rapid prototyping under tight deadlines |
| 📜 Certification | Tata Data Visualization: Empowering Business with Effective Insights (Forage, Jan 2026) |
| 🎓 CGPA | 9.80 / 10.0 through Semester 4 |