Use case
Game Economy Fraud Detection Demo
See how graph queries uncover suspicious trading patterns in a Roblox-style virtual economy. This PuppyGraph demo compares SQL joins with graph traversal and explores three fraud detection scenarios: mule networks, wash trading, and automated sniping.
We connect player and transaction data from a MySQL database, model players as nodes and trades as edges, and use PuppyGraph’s built-in AI assistant to create a graph schema and investigate activity in natural language.
PuppyGraph queries your existing relational data as a graph without requiring an ETL pipeline into a separate graph database.
Queries in natural language
- Find potential mule accounts funneling high-value items into a central account at unusually low prices.
- Identify wash-trading patterns where the same asset circulates through a closed loop of players
- Investigate rapid transaction sequences that may indicate automated bots or sniping scripts.
