Cybersecurity · Paris
AI agent risk assessment
- Problem
- Dozens of SaaS applications per company, each with a manual security review: slow, incomplete and out of date by the next version.
- Solution
- Scoping, design and delivery to production, as Head of Product, of a system of AI agents that assesses every SaaS application in the company: security posture, data handling, known breaches and vulnerabilities.
- Result
- An explained trust score for each application, and security recommendations generated automatically.
- gpt-oss
- Amazon Bedrock