Full-stack AI engineer focused on building agentic systems.

Currently working on multi-agent orchestration, memory, and RAG.

I work across Python (FastAPI, LangGraph) and TypeScript (Node.js, React, Next.js), and ship everything to prod.

Currently at 100xSchool Super30 (Greater Noida, India) — intensive full-stack bootcamp focused on AI and Web3.

Python FastAPI LangGraph Next.js React TypeScript PostgreSQL E2B Cloudflare R2 Docker Terraform Tailwind CSS

Describe what you want in plain English → a 2-agent LangGraph system (Sonnet 4.5 for builds, o4-mini for edits) generates production-ready React apps in isolated E2B sandboxes. Guardrail classifies intent, prompt enhancer expands vague ideas, assembler merges with a locked base template, and a validation loop auto-fixes build errors. Iterative chat refinement, live preview, one-click deploy to Cloudflare Pages. BYOK via OpenRouter.

TypeScript Next.js React Bun Express WebSocket Redis PostgreSQL TimescaleDB Prisma Docker Terraform Turborepo

7 microservices communicating via Redis Streams. In-memory liquidation engine checking every price tick across all open positions. BigInt arithmetic (10^8 scale) for zero floating-point errors. Live Binance price feeds, candlestick aggregation across 7 timeframes, snapshot-based crash recovery with event replay. Up to 100x leverage on BTC/ETH/SOL perpetuals.

🧠 re-collect — A memory layer for AI agents Python

Persistent, structured memory that lets AI agents recall context across sessions.

🛠️ craftsman — AI-powered coding tool Python

A developer tool that leverages LLMs to assist with code generation, editing, and reasoning across codebases.

Agentic AI Multi-agent Orchestration RAG Memory Systems Tool Use Prompt Engineering LangChain LangGraph Pydantic FastAPI SQLAlchemy Microservices WebSockets Redis Streams Node.js Express React Prisma Zod JWT CI/CD Machine Learning Deep Learning PyTorch

Languages

Frontend

Backend

AI & LLM

Data & Infra