I build production-grade AI applications — autonomous agents, RAG systems, and LLM-powered products — on top of a solid full-stack foundation (React/Next.js + Python). I specialize in taking AI products from zero to deployed, fast.
- 🤖 AI Agents — multi-step reasoning, tool/function calling, human-in-the-loop handoffs
- 🧠 RAG Pipelines — vector search, document retrieval, context-aware LLM responses
- 💬 LLM Integrations — OpenAI, Claude APIs, LangChain, LangGraph, streaming
- 🗃️ Memory Systems — short-term session memory + long-term persistent user context
- 📱 Full-Stack Apps — React/Next.js frontends, Django/FastAPI/Node backends, React Native mobile
AI / LLM
OpenAI API Claude API LangChain LangGraph RAG Vector DB Prompt Engineering MCP
Backend
Python FastAPI Django Node.js NestJS Express GraphQL REST WebSockets
Frontend
React.js Next.js TypeScript Zustand Redux Toolkit TanStack Query Apollo Client Tailwind CSS
Mobile
React Native Expo ML Kit Push Notifications React Navigation
Data & DevOps
PostgreSQL MongoDB ChromaDB Supabase Firebase Docker AWS CI/CD GitHub Actions
Autonomous agent that manages client conversations on Facebook/Instagram, qualifies leads with LLMs, and auto-books meetings synced to calendar availability.
- Multi-step agent with tool calling, conversation memory, and sales team handoff
- Built with OpenAI/Claude APIs + LangChain agent architecture
- Human-in-the-loop override system for quality control
Empathetic multi-turn AI counselor with safety-aware prompting and persistent session memory.
- LangChain-based agent with structured reasoning and guided decision support
- Long-term memory system that retains user context while enforcing privacy boundaries
- Thoughtfully designed prompt engineering for sensitive, high-stakes conversations
| Period | Role | Company |
|---|---|---|
| Apr 2025 – Present | Software Engineer (AI/Full-Stack) | Edutechs, Dhaka |
| Aug 2023 – Apr 2025 | Software Engineer (React + React Native) | W3kernel, Remote |
I'm open to remote Full-Stack AI Engineer roles and fast-paced startup environments.
If you're building an AI product and need someone who can handle both the LLM architecture and the full-stack implementation — let's talk.
