Software & AI Engineer · LLM Systems · AI Agents · RAG · Backend Architecture

Versatile Software & AI Engineer specializing in production-grade backend architectures, autonomous AI agents (LangGraph, n8n), and scalable microservices. I build stateful multi-agent pipelines, high-precision RAG systems, and deploy LLM-powered workflows to production.

🔭 Currently building: production AI agents with LangGraph & n8n workflow automation

💼 Open to AI/ML Engineering roles — remote or Dhaka-based

🎓 B.Sc. Computer Science & Engineering — Daffodil International University

🏆 NASA Space Apps Challenge 2024 — Regional Winner & Global Nominee (Top 50 worldwide)

🥉 NASA Space Apps Challenge 2023 — Regional 2nd Runner-Up

🖥️ ICPC Preliminary Contestant (2021, 2022)

AI Agents & LLM Engineering

RAG & Vector Search

Backend & Infrastructure

ML & Deep Learning

🔬 Autonomous Research Agent — Six coordinated agents (Planner → Search → Parser → Summarizer → Critic → Writer) that autonomously research any topic, with hybrid RAG (FAISS + BM25 + RRF) and real-time WebSocket streaming. FastAPI LangChain FAISS BM25 WebSockets Docker

⚖️ LexDraft — Legal Document AI — High-precision legal RAG system ingesting PDFs, scans, and handwritten notes via OpenCV + Tesseract OCR, with a semantic preference engine that learns from human edits. ChromaDB OpenCV Tesseract FastAPI Docker

🧠 MiniGPT from Scratch — Modern Llama-style decoder-only Transformer from scratch: custom BPE tokenizer, RoPE, SwiGLU, RMSNorm, Flash Attention, AMP mixed-precision training, and interactive Gradio demo. PyTorch RoPE SwiGLU Flash Attention Gradio

🤖 RAG Assistant API — Production FastAPI service with document ingestion, RAG-based chat, context memory, and tool calling. Powered by local embeddings. FastAPI LLM RAG ChromaDB Tool Calling

🏢 DealershipMS — Full-stack business management platform: FastAPI + SQLAlchemy 2.0 + PostgreSQL + React 19 + Redis, with JWT/RBAC auth, multi-language UI, and Docker Compose deployment. FastAPI PostgreSQL React Redis Docker