I'm a full-stack engineer at Stackular LLC (remote, Ellicott City MD), building production-grade web applications end-to-end — from database design through deployment. Most of my recent work is on Aurevo Fashion, a full-stack e-commerce platform demonstrating the complete SDLC: requirements → system design → API design → implementation → testing → deployment.
I pair traditional engineering discipline (typed APIs, real integration tests, CI/CD gates) with AI-assisted development workflows — using tools like Claude Code as a genuine engineering collaborator, not just autocomplete.
- 🔭 Currently building: Aurevo Fashion — Express/TypeScript backend + React 19 storefront, Supabase-backed
- 🌱 Exploring: AI agent tooling, knowledge-graph-backed codebase context, payment gateway integrations for the BD market
- 💬 Ask me about: React/TypeScript, Node.js/Express APIs, Postgres schema design, or AI-assisted engineering workflows
Planning & Design
Languages
Frontend
Backend
Database & ORM
Testing & QA
DevOps & CI/CD
Cloud & Hosting
Monitoring & Observability
AI-Assisted Engineering
| Project | Stack | Description |
|---|---|---|
| Aurevo.BE | Express · TypeScript · Drizzle ORM · Supabase | E-commerce REST API — modular monolith, JWT/JWKS auth, atomic stock accounting, AI shopping assistant |
| Aurevo.UI | React 19 · Vite · TanStack Query · Tailwind v4 | E-commerce storefront + admin panel, i18n (EN/বাংলা), no client-side Supabase SDK |
| Chatly | React · TypeScript · Firebase | Real-time chat app — 1:1 and group conversations |
| PH University Server | TypeScript · Express | Online courses platform backend |
| Doctor Portals | JavaScript | Dental doctor portal client |
Rebuilt Aurevo Fashion's chatbot from a bare Claude tool-use bot (no retrieval, no persistence, simulated streaming) into a full retrieval-augmented generation pipeline: semantic product search, policy/FAQ answers, and auth-gated order lookup — with real token streaming and conversation persistence. Full writeup: Aurevo.BE/docs/09-ai-chatbot-rag.md.
flowchart LR
subgraph ING["📥 Offline Ingestion"]
PROD[Products DB] --> EMB["Chunk + Embed (Voyage AI)"]
POL[Policy / FAQ docs] --> EMB
EMB --> KB[("kb_chunks — pgvector")]
end
subgraph RUN["💬 Runtime Chat"]
WID["Chat Widget (SSE)"] --> CS["Chat Service — Claude tool-use loop"]
CS -->|"search_knowledge"| KB
CS -->|"get_product_details"| PDB[(Live product DB)]
CS -->|"get_my_orders 🔒 auth-gated"| ORD[(Orders)]
CS --> OUT["Streamed reply + persisted conversation"]
end
ING --> RUN
Contribution snake above renders automatically via a scheduled GitHub Action — it'll populate a day after this repo is created.