https://maimoonamin.com/writing Results-driven Agentic AI Engineer specializing in architecting, orchestrating, and deploying scalable autonomous agent systems and LLM backends. Proficient in designing advanced multi-agent workflows using LangGraph, CrewAI, and AutoGen, supported by robust orchestration via LangChain and LlamaIndex. Experienced in optimizing local and cloud model inference (vLLM, Ollama) and building high-performance retrieval architectures utilizing vector databases (Qdrant, Pinecone, Milvus). Adept at delivering production-ready, context-aware AI solutions with stringent observability (Langfuse), safety guardrails (Guardrails AI), and resilient serving infrastructure (FastAPI, Docker, NixOS). https://maimoonamin.com/writing/orchestrating-langgraph-and-crewai-with-local-ollama Architectural blueprint from building AI-Learning-Career-Coach: using LangGraph for macro state transitions and CrewAI for collaborative multi-agent sub-tasks on local Ollama models. October 2025 https://maimoonamin.com/writing/asynchronous-multi-agent-critique-loops-django-celery-gemini Lessons from architecting Agentic-Review-Architecture: decoupling iterative multi-agent critique-refinement loops from HTTP request cycles using Django, Celery, and Redis. September 2025 https://maimoonamin.com/writing/high-performance-rag-qdrant-pinecone-milvus-nixos How to pair hybrid dense-sparse vector retrieval across Qdrant, Pinecone, Milvus, and PostgreSQL with reproducible NixOS and Docker serving pipelines. August 2025