Umar J.

AI agent developer and LLM engineer building production AI systems in Python: AI agents, RAG pipelines, LLM chatbots and voice agents, plus the FastAPI and Django backends they run on. 8 years of backend engineering, $200K+ earned across 73 projects.

Most of my work now is taking AI prototypes and making them reliable products, using OpenAI, Claude and Gemini. A demo that works for one user is easy. The real work is handling concurrent requests, keeping API costs under control, and failing gracefully when a model call times out or returns something unexpected.

What I build:

• AI agents and agent workflows that call tools and APIs, with retries, fallbacks and logging so you can see what the model did and why
• RAG pipelines: document ingestion, embeddings, retrieval and answer quality tuning
• LLM chatbots and voice agents, including streaming responses and live call automation
• LLM API integration: streaming, token limits, retry logic, cost control and observability
• Backends for AI products: REST APIs in FastAPI, Django REST Framework (DRF) or Flask
• PostgreSQL schema design and query optimization, Redis for caching and queues, Celery for background work, Docker for deployment

Recent projects:

• Voice agent platform for live call automation
• Modular AI finance tool
• AI case management system

How I work:

• I tell you upfront if something takes 3 weeks instead of 1
• I build for what happens when the happy path breaks, not just the demo
• Clear updates, and code your team can maintain after I'm done

Got an AI agent, LLM feature or Python backend that needs real engineering behind it? Send me a message and tell me what you're building.