AI Career
AI Developer and Data Scientist at NAVADA.
I help teams ship LLM systems that are reliable enough for production.
I design and implement LLM pipelines, RAG systems, and agent workflows end to end — from data ingestion and retrieval design to evaluation, deployment, and monitoring. My work sits at the intersection of software engineering and machine learning, with a strong focus on clear trade‑offs, observability, and keeping complexity under control.
LangGraph · LangChain · MCP
RAG architectures & vector search
Agents, tool use, integrations
Azure · OpenAI · Supabase
Recent work & focus areas
Lead and advise on AI initiatives for clients: from discovery and architecture to hands‑on
implementation. Typical work includes designing LLM‑powered assistants, evaluation loops,
and data pipelines that connect existing systems to new models.
Build retrieval‑augmented generation flows that turn internal knowledge (docs, tickets,
code, PDFs) into useful answers. Focus on chunking strategies, evaluation datasets, and
guardrails so systems stay robust as data and prompts change.
Experiment with multi‑agent patterns and the Model Context Protocol to connect LLMs to
tools safely — databases, APIs, code execution, and external services — while keeping
a clear map of dependencies and technical debt.
Turn real project experience into public notebooks, architecture guides, and visual maps
on LearnByAmir — so other builders can move faster without getting lost in the tool
landscape.