The U.S. Transportation Security Administration awarded MLtwist a $590K contract to label and process multimodal AI data. TSA uses the data to develop and deploy new threat detection models for checkpoint screening.
How it started
The award followed an aviation security pilot with Sandia National Laboratories, completed in May 2024. Sandia used MLtwist to build a reusable pipeline for labeling 3D screening scans and cut its data preparation time from eight weeks to three.
What MLtwist does for TSA
- Ingest. Take in raw screening scans and their metadata in a secure environment.
- Prepare. Clean and transform scans so they're ready to label.
- Label and verify. AI-assisted labeling, with people checking every item.
- Package. Deliver labeled data as JSON or in the native scan format, versioned and tracked at every step.
Why it matters
TSA's Open Architecture initiative uses shared, standards-based interfaces so many companies can build detection algorithms, user interfaces, and reporting tools. Those vendors all need the same thing first: consistent, well-labeled screening data with a clear record of where it came from.
"New AI capabilities applied to data processing, working in conjunction with human intelligence, deliver higher data quality with faster turnaround times. Thanks to the TSA's Open Architecture initiative, we are excited to be one of the many pieces involved in the future of aviation security."
— David Smith, founder and CEO, MLtwist
