We are looking for an AI specialist (AI Solutions Architect / Multi-Agent Engineer) to assemble and support a custom multi-agent system for legal tasks (Montenegro: corporate law/real estate/taxes/disputes/criminal law).
Need: to assemble a “council” of 5–6 different AIs (parallel responses) + a main Judge-AI that synthesizes one final, verified output; set up the loading of large arrays of materials (laws/regulations/case law, PDF/scans/documents), create a case database with long-term memory (knowledge base / RAG), implement strict verification (citations, links to sources, conflict detection).
Work format: long-term. The specialist will be a permanent assistant — maintaining quality and accuracy, improving prompts/agents/pipelines, testing new models and tools, tracking quality metrics, updating the system as changes occur.
Requirements: experience with multi-agent/orchestration (AutoGen/CrewAI/LangGraph/similar), RAG/vector databases, API integrations, quality assessment (eval), data security and privacy. Experience with legal projects and sources of law is welcome.
If interested — please send a brief portfolio/cases