Hands-on with LangChain/LangGraph, AutoGen, or CrewAICore agent patterns: tool use, memory, multi-step reasoning, output validationLLM APIs (OpenAI, Anthropic, Gemini, or open-source); structured prompt engineering
RAG & VECTOR INFRASTRUCTURE
End-to-end RAG pipelines: ingestion, chunking, embedding,retrieval evaluation
Familiarity with at least one vector database; ability to diagnose & improve retrieval quality
Hands-on with UiPath, Automation Anywhere, or Power AutomateBot workflows with exception handling, logging & integration with agent layersENGINEERING
Strong Python; cloud AI services; APIs, data pipelines & event-driven systems