- 🛠️ I build AI-powered web apps end to end: Next.js and React frontends, Node.js and FastAPI backends, and the LLM layer (tool calling, web search, RAG).
- 🤖 Currently building Devoploy, a GitHub App agent that reads failed CI runs, proposes a fix and opens a pull request.
- 🏭 Software Development Intern at ASA Industries, where I replaced a 141-column Excel workflow with a production ERP.
- 🎓 B.Tech ECE at SRM University (2028), looking for AI and full-stack internships.
AI & agentic AI: LLM integration · tool-calling agents · RAG with pgvector · structured outputs · classical ML
| Languages | |
| Frontend | |
| Backend | |
| Databases | |
| Tools | |
| IoT & hardware |   |
|
ContentSpy AI Paste a competitor's URL and get an AI-written intelligence report: SEO, content gaps and a market-entry plan, from an LLM with web search.   Live Demo | Code |
Pre-Delinquency Risk Intelligence Early-warning credit-risk scoring: 16 behavioural signals, a Random Forest, a FastAPI + MongoDB API and a React dashboard (simulated data).   Live Demo | Code |
|
Project Zenith team project · ASTRALWEB'26 AI-powered digital observatory with live ISS, satellite and planet data in a 3D interface. I built the Next.js frontend, 3D scenes and deployment.  Live Demo | Code |
Devoploy in progress A GitHub App agent that parses CI failures, proposes a fix with an LLM and opens a pull request; documentation search uses RAG over pgvector.  Code |
|
AI Algorithm Tutor React Native app that animates A* and alpha-beta pruning step by step, explained by a local LLM through Ollama.  Code |
CG Power: Equity Research Initiating-coverage report with an Excel model (DCF, reverse DCF, comps) and a Python script that checks every output.   Read it | Code |
