Andreas Kollegger builds multi-agent AI systems for developer relations at Neo4j. He focuses on making generative AI more reliable by grounding it in graph data structures.
His work involves giving AI agents better tools for planning, remembering, and retrieving information. He advocates for incremental adoption of Agentic GraphRAG, showing developers how connecting multi-agent systems with graph-memory can fix edge cases without requiring an immediate, complete architecture overhaul.