In addition to the subprocessors listed above who support the operation of our services, Deepgram may occasionally engage third-party speech recognition providers, including industry peers, to evaluate the performance and accuracy of our models. This comparative analysis is conducted under strict confidentiality and data protection controls.
Customers who do not wish to have their data used for model improvement or benchmarking purposes may opt out by:

• Opting out of participation in our Model Improvement Partnership Program, or

• Executing a Data Processing Agreement (DPA) with Deepgram that explicitly limits processing to core service delivery (please contact security@deepgram.com for details).