Customer Success Brain

Executive Summary
Memory Timeline
Knowledge Graph
Agent Workflow

Inspiration

Customer Success teams working on complex enterprise AI deployments face a familiar problem: critical customer context is scattered across meeting notes, support tickets, architecture documents, evaluations, and action trackers.

A project can look healthy in one document while newer evidence tells a completely different story. Important decisions get buried, ownership becomes unclear, and teams spend hours reconstructing what happened instead of deciding what to do next.

I wanted to build something that goes beyond answering questions about documents: a shared organizational memory that understands how customer context evolves and helps teams act on it.

What it does

Customer Success Brain is an AI-powered customer intelligence application that transforms fragmented project knowledge into evidence-backed launch decisions and actionable follow-ups.

Using a synthetic healthcare customer, Acme Health, the application demonstrates a realistic enterprise AI deployment approaching a critical production launch.

The system can:

Assess launch readiness: Evaluate five mandatory launch gates against documented requirements, including privacy, clinical accuracy, escalation, performance, and audit controls.
Remember new information: Store customer updates persistently in HydraDB and retrieve them alongside historical documents.
Reconcile conflicting evidence: Distinguish an older 97% prototype evaluation from a newer approved evaluation showing only 91% accuracy.
Explore relationships: Visualize knowledge-graph connections between stakeholders, risks