I build AI-powered products, scalable web applications, and real-time systems — from polished React interfaces to backend architecture, AI integrations, and systems-level engineering.
Next.js · React · TypeScript · Node.js · PostgreSQL · AI/RAG · Rust
I'm a full-stack developer based in Dhaka, Bangladesh, focused on building products that combine strong user experiences with reliable engineering.
My work spans modern web development, AI/LLM integration, real-time applications, and systems-level software.
- 🚀 Build production-focused applications from UI to backend
- 🤖 Work with LLMs, RAG pipelines, vector search, and AI APIs
- ⚡ Build real-time experiences involving audio, transcription, and streaming
- 🦀 Explore Rust and systems programming for performance-critical applications
- 🎨 Care deeply about clean UI/UX, accessibility, and interaction details
- 🧠 Continuously improving architecture, performance, and developer experience
| Area | Focus |
|---|---|
| Product Engineering | SaaS platforms, dashboards, marketplaces, internal tools |
| Frontend | React, Next.js, TypeScript, Tailwind CSS |
| Backend | Node.js, Express, PostgreSQL, Supabase, APIs |
| AI Engineering | RAG, embeddings, vector search, LLM integrations |
| Real-Time Systems | WebSockets, streaming, audio processing, transcription |
| Systems Engineering | Rust, Tauri, native desktop applications |
A cross-platform desktop application for real-time microphone and system-audio transcription.
Built with Rust + Tauri + React, the application captures multiple audio sources, processes them through a Rust audio pipeline, and supports both local and cloud-based speech recognition.
Highlights
- 🎤 Simultaneous microphone + system audio capture
- ⚡ Real-time speech-to-text
- 📴 Offline transcription using
whisper.cpp - ☁️ Cloud transcription using Groq
- 🦀 Rust-based audio processing
- 🪟 WASAPI loopback for system audio
- 🔄 Runtime switching between STT engines
Stack: Rust · Tauri · React · TypeScript · CPAL · Whisper.cpp · Groq
An AI-powered platform designed to help Bangladeshi students understand university admission requirements and discover suitable institutions.
The system combines structured university data, semantic search, RAG, and LLM-based responses to provide contextual admission guidance.
Highlights
- 🔎 Semantic search over university information
- 🤖 AI-powered admission assistance
- 🎯 Eligibility-based guidance
- 🧠 RAG-powered contextual answers
- 🗃️ Vector search with Qdrant
- 🔌 Multi-provider LLM architecture
Stack: Next.js · TypeScript · Qdrant · Drizzle · PostgreSQL · Ollama · OpenAI · Anthropic
An internal platform designed to streamline agency workflows around lead management, website auditing, and AI-assisted outreach.
Highlights
- Lead import and management
- Pipeline tracking
- Automated website audits
- AI-assisted outreach generation
- Structured client workflows
Stack: Next.js · TypeScript · TanStack Query · Zustand · Zod
A production-oriented blogging platform with role-based access and database-level authorization.
Highlights
- 👤 User / moderator / admin roles
- 🔐 PostgreSQL Row Level Security
- 📝 Content management workflow
- ✅ Approval and moderation system
- ⚡ Next.js App Router architecture
Stack: Next.js · Supabase · PostgreSQL · RLS
→ View Repository · Live Website
- Building practical AI-powered products
- Improving RAG architecture and retrieval quality
- Exploring real-time AI and speech systems
- Deepening my knowledge of Rust and systems programming
- Building better full-stack product architectures
- Creating software with a strong balance between engineering quality and user experience
I'm interested in collaborating on AI products, developer tools, SaaS platforms, real-time systems, and ambitious web applications.

