Detect Coordinated Fraud Rings via Link Analysis at a Neobank
Overview
What this challenge is about.
You receive 90 days of account, login, and transaction data (around 1.2 million accounts, around 30 million events) plus a labeled set of 80 known fraud rings. Build a multi-relation graph (accounts + devices + IPs + counterparties), apply community detection (Louvain or Leiden) and ring-suspicion features (component density, recency of joins, money-flow circularity), and produce a top-100 ring-suspicion list. Success is precision-at-50 above 0.6 against the labeled set, with a clear investigator-facing rationale per ring.
The Brief
What you'll do, and what you'll demonstrate.
Surface coordinated synthetic-identity fraud rings via graph and link analysis, ranked for investigator triage at precision@50 above 0.6.
Earning criteria — what you'll demonstrate
- Construct multi-relation graphs from operational data
- Apply community detection at fintech scale and reason about resolution
- Engineer ring-level features that align with investigator intuition
- Communicate model output to a non-ML investigative audience
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Data Scientist
Fraud-ring detection with graphs is a high-leverage data-scientist specialization at any fintech or marketplace; this challenge demonstrates the full workflow.
This challenge sharpens
- graph-analysis
- link-analysis
- fraud-detection
Applied AI Scientist
Translating community detection into investigator-actionable rationale is the kind of last-mile work applied AI scientists own.
This challenge sharpens
- community-detection
- fraud-detection
- feature-engineering
Machine Learning Engineer
Productionizing graph features and scoring pipelines is increasingly a junior MLE responsibility on financial-crime teams.
This challenge sharpens
- python
- feature-engineering
- graph-analysis