Orchestrate Entity Resolution for Cross-Border KYC with CrewAI and Cursor
Siloed legal entity databases cause high false-positive rates and slow clearance in corporate cross-border payments. Build a multi-agent entity resolution system using CrewAI and Cursor to link disparate entity registries across global financial networks. Achieve 95% entity disambiguation recall on noisy multi-jurisdictional firm records.
What you are building
The core problem, expected build, and operating context for this challenge.
Deploy a collaborative CrewAI agent team to extract, match, and resolve legal entity graph data across international registries.
How work is evaluated
Measure accuracy in resolving identical physical entities registered under differing structural names and jurisdictions.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
Correct Match Identification
Assert matches are correctly identified despite string variations.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Recall
Recall metric on corporate entity record matching test dataset. • target: 0.95 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Design multi-agent role delegation using CrewAI for KYC entity matching
Utilize Cursor AI features to generate scalable graph query execution pipelines
Link corporate records across heterogeneous sources using fuzzy matching and Graph Neural Networks
Generate consolidated Beneficial Ownership graphs for institutional counterparties
Reference links and supporting material
Synthetic dataset of 1,000 paired global corporate entity records with variations in tax format, corporate suffixes, and localized spellings.
Challenge setup
Measure accuracy in resolving identical physical entities registered under differing structural names and jurisdictions.
Challenge input
JSON containing record_A and record_B with name, address, tax_id, and jurisdiction.
CrewAI
Allows modular creation of specialized agents for collection, matching, and validation.
Cursor
Accelerates agent development and prompt iteration workflows.
Evaluated output
JSON containing is_match (boolean), confidence_score (float), and resolution_graph.
- Assert matches are correctly identified despite string variations.
- Recall metric on corporate entity record matching test dataset. • target: 0.95 • range: 0-1
- Benchmark: EntityMatchBench
- Recall target: 0.95
- 1 public reference case
- Python execution harness
- Python 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
- CrewAI · Required
- Cursor · Optional
- crewAI · Optional
- Cursor · Optional
Evaluation contract
- Correct Match Identification · Weight 1
- Recall · Weight 1
Metrics
- Recall · Target 0.95
Benchmarks
- EntityMatchBench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON containing record_A and record_B with name, address, tax_id, and jurisdiction.
Output: JSON containing is_match (boolean), confidence_score (float), and resolution_graph.
Execution harness: Python
Evaluation tools
- CrewAI · Agent Orchestration Engine
- Cursor · AI Development Platform
Recipe freeze marker
No recipe freeze time is recorded.