video1.0<iframe src="https://www.loom.com/embed/89ce38fd569b441085c162e6161fbe36" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/89ce38fd569b441085c162e6161fbe36-32d35f31a350cbf9.gif204.019This Loom demonstrates an engineering AI suite that coordinates retrieval and knowledge-graph queries to produce grounded answers. Mali shows three Python services that use gRPC internally and REST at the edge, with an orchestrator running a react loop where a planner selects tools, gathers evidence, and synthesizes responses. It covers a Neo4j knowledge graph with about 21 nodes and 25-ish relationships, plus retrieval via a Quadrant Cloud vector store. A demo asks which SILD requirements apply to the battery pack and which norms back them, showing traces and full audits with graph results, vector-store chunks, and the LLM-generated output.