AI/ML Engineer · Automation Systems · Fintech
Building AI systems that automate workflows, reason over domain knowledge, and process real-time data. Focused on LLM applications, RAG pipelines, automation infrastructure, and intelligent decision systems using FastAPI, Next.js, and modern AI tooling. Completed: LLM Post-Training Intern @ Ethara AI Currently: MCA graduating June 2026 · Open to ML engineering roles (remote/hybrid)
| ⭐ 40+ GitHub stars | CryptoMarket Regime Classifier |
| 📦 350+ downloads | Hugging Face training dataset |
| 📬 100+ emails/day | Shelly lead automation pipeline |
| 🧠 LLM Post-Training | RLHF evaluation @ Ethara AI |
CryptoMarket Regime Classifier — ⭐ 44+ stars
End-to-end ML pipeline that classifies market regimes (Trend, Range, Squeeze, High Volatility) from multi-timeframe crypto data using HMMs and LSTMs.
- Processed 2M+ market data points · identified 6 distinct regimes
- 6-state HMM trained on 315K BTC observations · deployed via Streamlit
- Dataset published on Hugging Face with 200+ downloads
- Modular design for integration into strategy, position sizing, and risk management
- 🤖 Opsora — independent automation consultancy; designing lead systems, follow-ups, and scalable workflow pipelines for businesses
- ⚡ Shelly — end-to-end lead generation engine: Google Maps scraping → email discovery → bulk outreach (100+ targeted emails/day, 90% less manual work)
- ⚖️ Lexx — citation-grounded legal AI for Indian law, built with Next.js + FastAPI + RAG
- 📊 Dazai — market-regime-aware research system with backtesting and risk-aware experimentation
- Build AI systems that reason, retrieve, and automate manual workflows end-to-end
- Create market intelligence tools that bridge human psychology and machine precision
- Explore how rule-based logic + LLM reasoning can power autonomous decision-making
If you're building in AI, trading, or automation or just exploring ambitious ideas - let's connect:
"Talk is cheap. Show me the code." — Linus Torvalds