Skip to main content
Back to challenges
Challenge

LEI Ownership Audit and AML Queue Automation with Mastra AI and o3

Unclaimed Legal Entity Identifier portfolios create systemic blind spots in institutional sanctions enforcement and ultimate beneficial ownership checks. You will construct a stateful workflow agent that extracts entity relationship trees and updates compliance queues, eliminating unassigned risk exposure across corporate structure datasets.

CybersecurityHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build a stateful workflow automation pipeline using Mastra AI in TypeScript powered by o3 reasoning to traverse LEI ownership structures.

Delivery guide

How work is evaluated

Evaluation

Evaluates Mastra AI step workflow accuracy in identifying root parent entities in multi-tier corporate hierarchies.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

root_entity_accuracy

Validates root LEI correctly identified

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

traversal_accuracy

Accuracy of ownership depth calculation • target: 1 • range: 0.95-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Orchestrate Mastra AI workflows and step graphs in TypeScript

  • Integrate OpenAI o3 reasoning model for complex graph traversal logic

  • Uncover Ultimate Beneficial Ownership (UBO) chains in orphan LEI portfolios

  • Automate investigator queue assignment for high-risk corporate entity clusters

Resources and assets

Reference links and supporting material

Dataset notes

GLEIF corporate structure dataset subset containing 1,500 complex multi-jurisdictional entity relationship trees.

Challenge setup

Evaluates Mastra AI step workflow accuracy in identifying root parent entities in multi-tier corporate hierarchies.

Challenge input

JSON representing nested corporate entity relations

Mastra AI

TypeScript framework for stateful agent workflows and RAG.

o3

Advanced reasoning engine for complex graph and entity reasoning.

Evaluated output

JSON containing root_parent_lei, ultimate_beneficial_owner, and depth

Evaluation checks
  • Validates root LEI correctly identified
  • Accuracy of ownership depth calculation • target: 1 • range: 0.95-1
Configured evaluation evidence
  • Benchmark: GraphAML-Bench
  • Traversal Accuracy target: 1
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • OpenAI o3 · Optional
  • Mastra AI · Required
Policy Serving
  • o3 · Optional
Orchestration
  • Mastra AI · Required

Evaluation contract

  • root_entity_accuracy · Weight 1
  • traversal_accuracy · Weight 1

Metrics

  • traversal_accuracy · Target 1

Benchmarks

  • GraphAML-Bench
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON representing nested corporate entity relations

    Output: JSON containing root_parent_lei, ultimate_beneficial_owner, and depth

    Execution harness: JavaScript

Evaluation tools

  • Mastra AI · agent_framework
  • o3 · llm_engine

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about LEI Ownership Audit and AML Queue Automation with Mastra AI and o3