I build end-to-end AI products β not prototypes, not proof-of-concepts. Systems that get deployed, used, and validated by real people. My work spans agentic AI, RAG architectures, fine-tuned language models, and multi-agent orchestration, always shipped to production on Hugging Face Spaces, Render, or Railway.
Philosophy: "Ship it. Get feedback from the people who matter. Then make it better."
AI & Orchestration: LangChain Β· LangGraph Β· LlamaIndex Β· Unsloth Β· FastMCP Β· Gemini 2.0 Β· DeepAgents SDK
| Project | Description | Live Link | Code |
|---|---|---|---|
| π notYC | AI-powered YC application reviewer. Validated by Garry Tan. | π Live | N/A |
| π₯ HCP-CRM | Multi-agent hospital workflow automation. | N/A | π» Repo |
| π Security SLM | Fine-tuned Llama 3.2 (1B) for security auditing. | π Live | π» Repo |
| π§ Sovereign Agent | System 1/2 reasoning agent with live web browsing. | π Live | π» Repo |
| π° Ledger AI | Multi-agent finance platform with PII guardrails. | π Live | π» Repo |
| π³ CC Optimizer | RL-based credit card simulation (Meta AI Finalist). | π Live | π» Repo |
| Hackathon | Outcome | Project Link |
|---|---|---|
| Meta AI India | Phase 2 Finalist | π Project |
| Aptos Blockchain | Recognition Award | π Project |
Self-updating ecosystems engineered to evolve daily via CI/CD pipelines.
| Lab | Focus | Repository | Live Interface |
|---|---|---|---|
| AI Academy | LLMs, RAG & Agents | π Repo | π Open |
| DSA University | SDE-2 Logic & Algorithms | π Repo | π Open |
"Ship it. Get feedback. Make it better."