Requirements
### Requirements
- 1+ year of production AI/LLM systems development (not academic pet projects)
- Strong Python skills
- Multi-agent orchestration / LLM-orchestration: agent frameworks (LangGraph-class), tool/function calling, structured output, guardrails
- RAG end-to-end: embeddings, chunking strategies, vector databases (Pinecone / pgvector / Qdrant), citation tracing, hallucination control
- Databases: SQL (PostgreSQL) + vector databases
- Data modeling, queries, understanding of data layer design
- Workflow automation: n8n (or similar) on real cases
- Basic LLM-ops: evals, observability (Langfuse-class), cost and latency control, fallback model chains
- Integrations: REST / Webhooks, working with external platform APIs and CRMs
- Practical experience with Claude (Anthropic) and/or OpenAI APIs
- Understanding of MCP (Model Context Protocol)
- English B1+