HorizonDB AI-Native Fraud Detection
Build an AI-native fraud detection system on HorizonDB.
- Beginner
- 6 hours
One day, guided
A single facilitated day. Every challenge validates automatically, so the room self-paces and the leaderboard settles itself. No jury needed.
Available on request · we reply within one business day
Overview
Provision a zone-redundant cluster, index transaction embeddings with DiskANN, detect fraud rings with Apache AGE graph queries, run SQL-native LLM calls for hybrid search, and expose a fraud-check tool to a Microsoft Foundry agent.
What participants work through
- HorizonDB Cluster Provisioning & HA - zone-redundant; forced failover test
- AI Pipeline & Vector Embeddings - DiskANN-indexed transaction embeddings
- Graph Reasoning with Apache AGE - ring fraud detection via Cypher queries
- AI Functions & Hybrid Search - SQL-native LLM calls; SAR narrative generation
- Data API Builder MCP Exposure - fraud-check tool in Microsoft Foundry agent
Run this with your teams
Environments, coaching, judging and reporting, all handled.