AI Application Risk Classifier Partner built app
Get quick answers to common questions about the AI Application Risk Classifier, including agent infrastructure, supported frameworks, and the models used for analysis.
Where's the Tavro agent based and how does the app communicate with it?
The Tavro Agent runs on AWS. Communication between Atlan and the Tavro API happens over HTTPS (TLS 1.2 or higher), ensuring that all content in transit is encrypted.
What regulatory frameworks are supported for analysis?
The Tavro AI Agent's analysis aligns with the EU AI Act, which defines AI system obligations based on risk levels related to safety, fundamental rights, and ethical standards.
What models does the Tavro AI agent use?
The agent is powered by OpenAI's Large Language Models (LLMs). These models interpret and analyze AI application metadata and map contextual meaning to regulatory categories for risk assessment.
See also
- Understanding metadata risk classification: Understand the risk classification framework and how the AI Application Risk Classifier evaluates metadata.
- Configure risk metadata: Set up the workflow with configuration parameters and view output specifications.