About Me: I am a recently graduated AI Engineer specializing in moving LLMs out of research notebooks and into robust, production-ready backend services. I focus on stateful multi-agent architectures, dynamic RAG pipelines, and strict structured outputs. Available for remote, long-term B2B contract roles starting immediately.
What I Build:
-
Agentic Orchestration: Designing stateful, multi-agent workflows using LangGraph and LangChain (e.g., Planner, Researcher, Critic loops).
-
Backend & APIs: Wrapping AI logic in robust FastAPI endpoints, enforcing strict data contracts with Pydantic, and fully containerizing the pipeline with Docker.
-
LLMs & Fine-Tuning: Leveraging frontier APIs (Claude, OpenAI) and fine-tuning open-weight models locally (like Qwen via LoRA/Unsloth) to optimize inference costs and reduce latency.
-
Production RAG: Moving beyond static vector databases by integrating dynamic tool calling (e.g., live web retrieval via Tavily API) with automated citation tracking and programmatic evaluation harnesses to prevent hallucinations.
Recent Highlight: Architected a fully containerized 6-agent research state machine that autonomously retrieves live data, synthesizes structured reports, and evaluates its own output via a reflection loop before returning the final payload.
Github: ebrahimzaher (Ebrahim Zaher)
Availability: Available for remote B2B contracts. Let’s connect if you need an engineer who focuses on pragmatism, cost-efficiency, and systems that actually work in production.