Senior Backend Engineer
Architecting high-performance distributed systems in Go & Python. 7+ years of experience in AI/RAG infrastructure, real-time systems, and cloud-native solutions.
Senior Backend Engineer with 7+ years of experience architecting high-performance distributed systems in Go and Python. Deep expertise in AI/RAG infrastructure, including vector databases (Milvus), custom embeddings, and low-latency real-time pipelines (WebSockets, gRPC).
Proven track record of productionizing ML platforms and handling 5,000+ concurrent connections. Based in Iraq and open to fully remote roles or relocation to the EU/Japan, with employer visa/work-permit documentation support where available.
IQ Labs
Built a real-time vehicle tracking system (buses/taxis) using Go, WebSockets, and PostGIS, handling 5,000+ concurrent connections with <200ms P95 latency and zero data loss under peak load.
Designed and implemented a production ISP billing system with RADIUS integration, enabling real-time provisioning and automated subscription enforcement, eliminating manual workflows.
Engineered and productionized an ML/RAG platform using Milvus vector database, custom embeddings, and re-ranking models — reduced response latency from ~2s to <400ms, introduced streaming SSE responses, replaced REST with gRPC between Go and Python services.
Provisioned and managed AWS EC2 GPU instances to host ML models, significantly accelerating inference.
Built a vendor-agnostic voucher and billing middleware integrating Microsoft Dynamics, reducing reconciliation workload by ~70% and cutting vendor onboarding from weeks to hours.
Delivered multiple production systems: nationwide IoT telemetry ingestion pipeline (air quality sensors), automated KYC verification backend, and full backend for Kurdim (language learning platform).
Neofinity & Client Projects
Managed end-to-end backend development for various client projects, transitioning legacy PHP applications to modern Python and Go architectures.
Migrated the Kurdoc platform (later rebranded as Gloctor) from PHP to Python, then rewrote core services in Go to resolve performance bottlenecks and improve API response times.
Built custom POS and retail systems from the ground up, implementing real-time inventory tracking, IMEI-based device management, and financial reporting integrations.
Go • Python
REST APIs • gRPC • GraphQL • WebSockets
PostgreSQL • PostGIS • Redis • Milvus
Distributed Systems • Microservices • Real-time Streaming
Linode/Akamai • AWS EC2/GPU • Docker • Terraform • CI/CD • Prometheus • Grafana
Go, WebSockets, PostGIS, Redis
Problem: Needed low-latency location updates for buses and taxis under commuter peak load.
Owned: Designed the ingestion pipeline, WebSocket fanout, geospatial queries, and reliability path.
Result: Handled 5,000+ concurrent connections with <200ms P95 latency and zero peak-load data loss.
Go, RADIUS, PostgreSQL, integrations
Problem: Manual subscription changes slowed provisioning and made enforcement inconsistent.
Owned: Built real-time billing, provisioning, and subscription enforcement workflows.
Result: Removed manual operational steps and enabled automated subscriber state changes.
Milvus, gRPC, Go, Python, AWS EC2 GPU
Problem: Retrieval and generation latency was too high for production user experience.
Owned: Productionized vector search, custom embeddings, re-ranking, streaming SSE, and gRPC service boundaries.
Result: Reduced response latency from ~2s to <400ms while supporting streaming responses.
Created go-arabic-reshaper, a high-performance Go library for Unicode-accurate Arabic text rendering in PDFs and images.
Professional Certificate — June 2023. Focus: API Design, Database Optimization, and Cloud Deployment.
Komar University of Science and Technology
Graduated January 2024
Based in Iraq and open to fully remote roles or relocation to the EU/Japan. Willing to support relocation logistics where employer visa/work-permit documentation is available.