Senior Backend Engineer with 6+ years building secure, high-availability systems in Go. Currently at go-e Labs, rearchitecting a high-load EV-charging platform for autoscaling and real visibility. From event-driven microservices and Kafka streaming to Kubernetes and deep observability, I build backends that don't break.
Re-architected a monolithic Node.js proxy into focused Go microservices with an event-driven Redis design, migrating from PM2 to Kubernetes on OVH with autoscaling. Cut p99 latency ~55% and memory ~40% while supporting ~2.5× more concurrent charger connections.
Rebuilt the charging-session service with hot-path preloading and an event-driven pipeline, cutting session generation from minutes to under 10s and export to ~30× faster, with ~85% lower p95 and far fewer support tickets.
Introduced end-to-end observability to a previously blind platform, standing up SigNoz (OpenTelemetry) and evaluating Datadog, giving the team real production visibility and faster time-to-detect for the first time.
Connected Hetzner and OVH over a WireGuard tunnel for secure ClickHouse access, and automated deployments with ArgoCD GitOps on Kubernetes (deploys from ~15 min to ~2 min).
Architected 15+ microservices on Kubernetes serving 500K+ users. Implemented Kafka event-driven architecture processing 1M+ events daily. Set up Prometheus/Grafana monitoring. Designed PostgreSQL schemas with optimized indexing. 99.99% availability.
Integrated Stripe and local payment providers with secure OAuth2 flows. Built REST APIs serving 200K+ customers. Redis caching reduced database load by 60%. PCI-DSS compliant with 85% test coverage.
Developed scalable Go services on AWS with Terraform. Built data pipelines for analytics dashboards. Ensured GDPR compliance with strict security controls.
Built Python/FastAPI backend for intelligent document querying system with secure API integrations.
Built Go microservices powering a copy-trading engine with Kafka event streaming and sub-millisecond Redis caching. Profiled with pprof to eliminate bottlenecks. Managed infra with Terraform and Docker. 99.9% uptime.
Built a secure Python/FastAPI backend with LLM integration and RAG pipelines for intelligent document querying, with data security and access control. 85% automated test coverage.
Designed a GDPR-compliant backend with RBAC, OIDC, comprehensive audit logging, and encryption for sensitive data.
Built backend services processing real-time sensor data streams, with time-series data models optimized for query performance.
Maintained cloud-native Go services on Kubernetes handling 300M+ daily requests. Optimized MySQL and Redis queries for high-traffic loads. Improved resilience through circuit breakers and monitoring. Performance optimizations improved response times by 25%.
Built Go/C++ applications handling 50K+ concurrent connections with focus on resilience. Integrated ML models into production services. Managed deployments with Ansible on Linux.
Optimized PostgreSQL queries reducing API response times by 60%. Implemented robust testing strategy improving release quality.
Umar delivered exceptional backend work on our copy trading platform. His deep expertise in Go and gRPC made our real-time trading engine incredibly fast and reliable. Communication was outstanding throughout the project.
Working with Umar on our healthcare platform was a great experience. He built GDPR-compliant services with meticulous attention to security. His understanding of RBAC and OIDC saved us months of development time.
Umar is one of the strongest Go engineers I've worked with. His ability to design resilient microservice architectures and lead backend teams through complex fintech projects is remarkable. He consistently raised the bar for our entire engineering org.
Umar stood out for his ownership mentality and deep technical skills. He didn't just write code — he thought about the bigger picture: reliability, security, and long-term maintainability. A valuable asset to any engineering team.
Collaborating with Umar at IBM was a pleasure. He maintained services handling hundreds of millions of daily requests with incredible discipline. His approach to error handling and circuit breakers significantly improved our API resilience.
Umar was the best mentor I could ask for. His code reviews were thorough and educational — he didn't just point out issues, he explained the "why" behind every decision. I grew more as an engineer in a year under Umar than in the previous three.
High-throughput order matching engine built for maximum TPS. Features a custom Write-Ahead Log (WAL) with inner/outer log separation, snapshotter for state recovery, Red-Black tree for order book management, and RCU-style deletion for lock-free reads. Hot-path is allocation-free. gRPC API with Kafka integration for order broadcasting.
REST API that generates creative brand names using OpenAI based on a business idea, then checks domain availability via the GoDaddy API in real-time. Clean architecture with separate API, service, and config layers. Dockerized for deployment. Live at brandsdigger.com.
Microservice-based cryptocurrency price aggregator. Data service fetches real-time prices from multiple exchanges, publishes via RabbitMQ, and a separate WebSocket service pushes live updates to clients. Designed for horizontal scaling. Work in progress.
Open source multi-signature signing application for Algorand blockchain. Enables devs to easily spin up multi-sig signing workflows for secure on-chain transactions.
Peer-to-peer networking library in Go for cryptocurrency networks. Implements node discovery, message propagation, and connection management over TCP/IP.
Desktop workbench for TCP client/server communication testing. Built with C++ and Qt framework for debugging network protocols.
A deep-dive into the pitfalls of ORM abstraction in Go backends and how to avoid them at scale.
🎤 Speaker at local tech meetups on Go best practices and microservice architecture
🎓 B.Sc. Computer Science — FAST-NUCES, Islamabad
Based in Germany, open to relocation (Munich and Berlin especially).
German work permit holder. Available from January 2027.