Context Graph is an AI-powered decision tracing system designed for financial institutions. It captures, stores, and analyzes the reasoning behind every decision made by AI agents and human operators.
Using a knowledge graph powered by Neo4j, the system maintains full context and provenance for decisions, enabling transparency, auditability, and continuous improvement.
Ask the AI Assistant
Use natural language to search for customers, review decisions, or analyze patterns. The AI has access to the full context graph.
Explore the Context Graph
Visualize entities and their relationships. Double-click nodes to expand and explore connected data. Click nodes to inspect properties.
Trace Decisions
Select any decision to see its full reasoning, causal chain, and similar precedents. Understand why decisions were made.
Find Patterns
Use graph algorithms to detect fraud patterns, find similar decisions, and identify influential precedents across the organization.
The context graph contains the following entity types and their relationships. This schema visualization shows how data is connected in the knowledge graph.
Failed to load schema. Please try again.
AI and human decisions with full reasoning traces
Customers and individuals in the system
Financial accounts owned by persons or organizations
Financial transactions between accounts
Companies, employers, and counterparties
Fraud alerts and compliance notifications
Customer support tickets and inquiries
Business rules and compliance policies
Internal staff who make or review decisions
Ask questions about customers, decisions, and policies
Welcome to Context Graph Demo
I can help you search for customers, analyze decisions, find similar precedents, and trace causal relationships. Try asking:
→ Should we approve a credit limit increase for Jessica Norris? She's requesting a $25,000 limit increase.
→ Search for customer John Walsh
→ A customer wants to make a $15,000 wire transfer. What policies apply and are there similar past decisions?
→ We need to override the trading limit for Katherine Miller. Find precedents for similar exceptions.
Visualize entities, decisions, and causal relationships
Inspect reasoning, precedents, and causal chains
Use the AI assistant to search for customers or decisions. Decision nodes will appear here when added to the graph.
No decisions in graph yet.