Build an M&A Technical Due Diligence Agent with OpenAI Agents SDK
Fintech acquisition due diligence often overlooks hidden AI technical debt, licensing risks, and model vulnerabilities. Build an OpenAI Agents SDK system to scan target codebase repositories, audit AI model lineage, and evaluate financial crime compliance gaps during M&A processes. Deliver an automated valuation risk assessment report with quantified technical risk scores.
What you are building
The core problem, expected build, and operating context for this challenge.
Build an automated due diligence agent using OpenAI Agents SDK to analyze software repositories, model weights provenance, and regulatory compliance risks during fintech acquisitions.
How work is evaluated
Evaluates accuracy of codebase scanning and licensing risk identification across fintech target code bases.
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.
risk_score_in_range
Checks if risk score is a decimal value between 0.0 and 1.0
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
license_detection_recall
Recall in detecting restrictive open source licenses • 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
Construct an agentic repository auditor using OpenAI Agents SDK tool invocation
Scan target open-source dependencies for restrictive licenses and vulnerability CVEs
Audit proprietary model training lineage for potential copyright and privacy compliance gaps
Generate aggregated M&A technical debt and risk valuation reports
Reference links and supporting material
Dataset of 150 simulated fintech dependency manifests with embedded software licenses and vulnerability signatures.
Challenge setup
Evaluates accuracy of codebase scanning and licensing risk identification across fintech target code bases.
Challenge input
JSON containing requirements_txt or package_json content
OpenAI Agents SDK
Standardized OpenAI agent framework providing clean function calls and handoffs.
Evaluated output
JSON with high_risk_licenses, cve_count, overall_risk_score, summary
- Checks if risk score is a decimal value between 0.0 and 1.0
- Recall in detecting restrictive open source licenses • target: 0.95 • range: 0-1
- Benchmark: M&A Technical Due Diligence Benchmark
- License Detection 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
- OpenAI · Required
- Azure OpenAI Service · Optional
- OpenAI o3 · Optional
Evaluation contract
- risk_score_in_range · Weight 1
- license_detection_recall · Weight 1
Metrics
- license_detection_recall · Target 0.95
Benchmarks
- M&A Technical Due Diligence Benchmark
- security
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON containing requirements_txt or package_json content
Output: JSON with high_risk_licenses, cve_count, overall_risk_score, summary
Execution harness: Python
Evaluation tools
- OpenAI Agents SDK · Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.