
    Rapid7 automates vulnerability risk scoring using Amazon SageMaker by creating an end-to-end ML pipeline that trains, validates, and deploys eight parallel machine learning models to predict CVSS vectors in near real-time.


<div>
<p>
This article details how Rapid7 used Amazon SageMaker to automate vulnerability risk scoring using machine learning pipelines for CVSS (Common Vulnerability Scoring System) vector prediction.
</p>
<ul>
<li>Developed an end-to-end automated ML pipeline for predicting CVSS vulnerability scores</li>
<li>Created eight parallel models to predict different CVSS vector metrics</li>
<li>Implemented continuous integration and deployment (CI/CD) for ML models</li>
<li>Used SageMaker Pipelines for orchestrating model training and deployment</li>
<li>Utilized SageMaker Model Registry to track and manage model versions</li>
<li>Deployed models using inference components for cost efficiency</li>
</ul>
<p>
The solution allows Rapid7 to automatically generate CVSS scores for new vulnerabilities, saving 2-3 days of manual work monthly and reducing compute costs by approximately 50%.
</p>
</div>


Related articles