Skip to content
Available for senior full stack roles

Md. Omar Faruk KhanFull Stack Developer & Systems Architect

Architecting high-performance web applications, scalable database systems, and microservice workflows.

  • Next.js
  • Node.js
  • PostgreSQL
  • MongoDB
  • Docker
  • Redis
Md. Omar Faruk Khan — Full Stack Developer & Systems Architect

Uptime

99.9%

Queue throughput

10k jobs/hr

p95 latency

< 200ms

Years building for production
4+
Records lost in a 1M+ row migration
0
Product teams shipped with end to end
2
Services running on self-managed VPS
10+
About

I build the parts of a product that have to survive scale.

Senior full stack developer with 4+ years shipping production systems end to end — from SSR-rendered Next.js frontends to Node.js services, queue-driven microservices and the Linux infrastructure they run on. I specialise in data-heavy platforms: schema design, migration at scale, caching strategy and the kind of query tuning that turns a 4-second page into a 200ms one.

Based in Dhaka, Bangladesh, working with teams anywhere. Most of my recent work sits where product engineering meets infrastructure — migrations, queues, caching layers and the observability to prove they behave.

End-to-end ownership

Interface, API, database and the box it runs on — one person accountable for the whole path.

Performance as a feature

Index strategy, Redis caching and SSR budgets treated as product requirements, not afterthoughts.

Systems that decouple

Queue-driven microservices with BullMQ so slow work never blocks a user's request.

Ship, then keep it up

Dockerised deploys, Nginx routing and GitHub Actions pipelines on self-managed Linux VPS.

Experience

A track record of shipping and keeping it running.

Click any role to expand the engineering detail behind it.

  • Architected high-performance web applications with React.js, Next.js, Node.js and Express.js.
  • Executed a large-scale data migration from MySQL to MongoDB with zero data loss.
  • Engineered scalable schemas in PostgreSQL and MongoDB with optimized query indexing.
  • Integrated BullMQ background queues to automate microservice workflows.
  • Enhanced real-time performance using Redis caching and WebSocket push notifications.
  • Managed Linux VPS, cloud infrastructure, Docker and Nginx configurations.
Next.jsNode.jsExpress.jsPostgreSQLMongoDBRedisBullMQDockerNginx
Tech Stack

The tools I reach for, grouped by the problem they solve.

Depth over breadth — every item here has shipped to production, not just a tutorial.

Languages & Frameworks

The application layer — typed, tested and server-rendered.

  • Node.js
  • Express.js
  • NestJS
  • TypeScript
  • React.js
  • Next.js

Databases & ORMs

Schema design, indexing strategy and caching that holds under load.

  • PostgreSQL
  • MongoDB
  • MySQL
  • Redis
  • Prisma
  • Mongoose

DevOps & Infrastructure

From a bare Linux box to an automated deploy pipeline.

  • AWS
  • Docker
  • Nginx
  • Vercel
  • VPS Management
  • GitHub Actions
  • GitLab

AI & Integrations

LLM workflows and third-party APIs wired into real products.

  • OpenAI API
  • GitHub Copilot
  • Gemini
  • Ollama (Local LLMs)
  • Stripe
  • Uber Direct API
Node.jsExpress.jsNestJSTypeScriptReact.jsNext.jsPostgreSQLMongoDBMySQLRedisPrismaMongooseAWSDockerNginxVercelVPS ManagementGitHub ActionsGitLabOpenAI APIGitHub CopilotGeminiOllama (Local LLMs)StripeUber Direct APINode.jsExpress.jsNestJSTypeScriptReact.jsNext.jsPostgreSQLMongoDBMySQLRedisPrismaMongooseAWSDockerNginxVercelVPS ManagementGitHub ActionsGitLabOpenAI APIGitHub CopilotGeminiOllama (Local LLMs)StripeUber Direct API
Projects

Systems built to carry real load.

Selected engineering work — the migrations, queues and AI layers behind the products I've shipped.

Database & Data Pipeline Engine preview
Data Engineering

Database & Data Pipeline Engine

High-performance MySQL → MongoDB migration system.

A resumable migration engine that streams relational tables into document collections in batches, mapping foreign keys to embedded and referenced shapes on the way. Checkpointing plus row-level checksums made a multi-million-row cutover repeatable — and verifiably lossless.

  • Batched, back-pressured streaming keeps memory flat regardless of table size
  • Checkpoint + resume on failure, so a dropped connection never restarts the run
  • Post-migration checksum reconciliation proving zero data loss
Node.jsMySQLMongoDBTypeScriptDocker
Asynchronous Task Queue Service preview
Distributed Systems

Asynchronous Task Queue Service

Microservices automated with BullMQ, Redis and a WebSocket push layer.

A job orchestration service that decouples slow work — invoicing, notifications, third-party sync — from the request cycle. Workers scale horizontally behind Redis, retries use exponential backoff with a dead-letter queue, and progress streams back to the browser over WebSockets.

  • Priority queues with exponential backoff retries and a dead-letter lane
  • Live job progress pushed to clients over WebSockets — no polling
  • Horizontally scalable workers, deployed as independent Docker services
BullMQRedisNode.jsWebSocketDockerNginx
AI-Powered Local LLM Integration preview
AI Engineering

AI-Powered Local LLM Integration

Workflow automation on Ollama with an OpenAI-compatible fallback.

A provider-agnostic AI layer that routes prompts to locally hosted models via Ollama and falls back to the OpenAI API when a task needs more capability. Keeps sensitive documents on-premise while still exposing streaming completions and structured tool output to the product.

  • Pluggable provider interface across Ollama, OpenAI and Gemini
  • Streaming responses with token-level SSE delivery to the client
  • Prompt + embedding cache in Redis to cut repeat inference cost
OllamaOpenAI APINext.jsRedisTypeScript
Real-Time Delivery Operations preview
Product Engineering

Real-Time Delivery Operations

Stripe billing and Uber Direct dispatch inside one operations console.

An operations dashboard where orders, courier dispatch and payment state stay in sync in real time. Uber Direct handles last-mile logistics, Stripe handles money, and a Redis-backed event fan-out keeps every connected dashboard on the same tick.

  • Uber Direct dispatch with live courier tracking on the order timeline
  • Stripe payments, webhooks and idempotent reconciliation
  • Redis pub/sub fan-out so every open dashboard updates instantly
Next.jsPostgreSQLPrismaStripeUber Direct APIRedis
Education

Formal grounding, sharpened in production.

B.Sc. in Computer Science & Engineering (CSE)

Bangladesh Institute of Science & Technology (BIST), Dhaka

Affiliated with National University of Bangladesh

Graduated 2021

Continued focus

  • Distributed systems & queueing
  • Relational + document data modelling
  • Linux server administration
  • Applied LLM integration
Contact

Have a system that needs building — or rescuing?

Tell me what you're working on. I reply to every message within a day or two.

Location

Uttar Badda, Dhaka, Bangladesh

Elsewhere