Still Measuring Initiative Impact Manually? How We Used Graph RAG + MCP to Make It Explorable
Measuring 'did that initiative actually work?' usually means manual SQL spelunking. We modeled initiatives × KPIs as a graph and let an LLM traverse it via MCP.