Case Study: Neural Guard achieves scalable AI detection workflows with ClearML

Neural Guard’s AI Workflow The Image of Efficiency

Neural Guard builds AI-based auto-detection solutions for security screening, using deep learning to identify high-risk items in CT and X-Ray imagery. The company needed a scalable way to manage large, constantly changing datasets and multiple object detection models across many machine and customer environments, while keeping costs down and minimizing manual DevOps and data engineering work. ClearML’s experiment management, MLOps, and data management platform was used as the core infrastructure for this effort.

Using ClearML, Neural Guard built an automated pipeline to ingest, analyze, label, QA, version, train, compare, and deploy models, with custom integrations through SDKs and APIs plus on-premise deployment support. The result was a highly scalable production workflow that helped Neural Guard better understand its data needs, reduce the amount of training data required, and avoid building a custom data management system from scratch. ClearML reportedly saved them decades of man-hours, reduced ongoing staffing and infrastructure costs, and accelerated time-to-market while supporting a best-in-class detection solution.

Neural Guard

Raviv Pavel

Chief Technology Officer