I’m a software engineer from Bangladesh passionate about building scalable applications, cloud-native infrastructure, and high-performance distributed systems.
My core expertise includes Go, TypeScript, Node.js, React, Kubernetes, Docker, AWS, CI/CD pipelines, microservices architecture, observability, and cloud automation. I enjoy designing systems that are reliable, cost-efficient, and production-ready.
Over the years, I’ve worked on projects ranging from AI-powered RAG platforms and fault-tolerant ticketing systems to adaptive video streaming infrastructure and intelligent Kubernetes autoscaling solutions.
🚀 A few things I love building:
- Cloud-native platforms powered by Kubernetes, Docker, and AWS
- Scalable backend systems with microservices and event-driven architecture
- AI-integrated applications using LLMs, RAG pipelines, and real-time streaming
- DevOps & automation workflows with Infrastructure as Code, CI/CD, and observability tooling
- Performance-focused distributed systems designed for scalability, resiliency, and real-world production workloads
I enjoy solving complex engineering challenges, optimizing infrastructure costs, and building systems that can reliably scale to serve thousands of users.
| Judge | Count | Profile Links |
|---|---|---|
| Codeforces | 250+ |  |
| LeetCode | 100+ |  |
| CodeChef | 150+ |  |
| HackerRank | 100 |  |
| Name | Description | Project Link | Video Link |
|---|---|---|---|
| NodeFleet | AWS K3s autoscaler reducing costs by 40-50% via Lambda, Prometheus metrics, and intelligent EC2 scaling with spot instances, Multi-AZ support, and predictive scaling. | Github | Video |
| Askly | An intelligent open source RAG engine powered by Google Gemini - chat with your PDFs, stream AI answers word-by-word via SSE, and fall back to live web search automatically. | Github | |
| Jatra | Fault-tolerant train ticketing system solving Bangladesh Railway's Eid traffic chaos. Built with microservices, K8s, Redis & OpenTelemetry. | Github | |
| VisionSync | Scalable cloud-native video streaming platform built on AWS. Features adaptive DASH/HLS streaming, real-time processing with ECS Fargate (70% Spot), MongoDB replica set, Redis caching, Socket.IO updates, and full IaC automation with Pulumi & Ansible. Cost-optimized multi-AZ architecture with CloudFront CDN. | Github | |
| SimplyDone | SimplyDone is a scalable To-Do app built with React.js (frontend) and Node.js (backend). Both are containerized with Docker and deployed on AWS EC2 instances with an Application Load Balancer (ALB) for high availability and fault tolerance. Pulumi is used as IaC | Github |
