M&A Merchant Portfolio Risk Due Diligence System with CrewAI & Devin
Payment company M&A deals often fail due to hidden merchant risk exposure and unvetted processing portfolios. Build a multi-agent audit system using CrewAI and Cognition Devin to ingest merchant portfolios, calculate chargeback risk scores, and generate risk reports. Achieve 95% precision on elevated merchant risk detection across cross-border processing datasets.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate specialized CrewAI role-playing agents paired with Cognition Devin to perform deep portfolio risk audits for fintech M&A transactions.
How work is evaluated
Evaluates crew performance on identifying high-risk merchant portfolios and precision of chargeback prediction summaries.
Shared data for this challenge
Review public datasets and any private uploads tied to your build.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
audit_schema_valid
Ensures output JSON contains all required risk assessment parameters.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
elevated_risk_precision
Precision of identifying truly high-risk merchant entities. • 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
Configure role-playing CrewAI agents for portfolio risk, regulatory compliance, and financial health auditing
Utilize Cognition Devin for autonomous code generation and dataset ingestion workflows
Build dynamic chargeback and fraud forecasting pipelines using portfolio metrics
Generate comprehensive executive risk summaries with structured action items
Reference links and supporting material
Synthetic M&A merchant dataset representing 50 payment processing accounts with varied chargeback rates, MCC codes, and cross-border volume.
How this agent runs
Evaluates crew performance on identifying high-risk merchant portfolios and precision of chargeback prediction summaries.
Challenge input
CSV/JSON containing merchant transaction history, chargeback ratios, and MCC codes
CrewAI
Framework for orchestrating role-playing autonomous agents.
Cognition Devin
Autonomous software agent for data pipeline code generation.
Evaluated output
JSON audit report with overall portfolio risk score and flagged high-risk merchant IDs
- Ensures output JSON contains all required risk assessment parameters.
- Precision of identifying truly high-risk merchant entities. • target: 0.95 • range: 0-1
- Benchmark: FintechRisk-MA-v1
- Elevated Risk Precision target: 0.95
- 1 public reference case
- Python execution harness
- Python sandbox
View technical recipe
Configured tools
- CrewAI · Required
- crewAI · Optional
- Cognition Devin · Optional
- Cognition Devin · Optional
- Cognition Devin · Optional
Evaluation contract
- audit_schema_valid · Weight 1
- elevated_risk_precision · Weight 1
Recipe state
This is a preview. The configuration can change before the evaluation recipe is locked.