DigitalOcean GenAI Platform lets you build GPU-powered AI agents with fully-managed deployment. Agents can use pre-built or custom foundation models, incorporate function and agent routes, and implement RAG pipelines with knowledge bases.
We do not store agent inputs or outputs on DigitalOcean infrastructure. Agent inputs and outputs for models hosted on DigitalOcean infrastructure (Llama, Mistral, DeepSeek) are stored in the local browser storage cache of the browser used to access the agent.
For agents using models from third-party model providers like Anthropic, the data is stored in accordance with that model provider’s policies. Reference your model provider’s policies for more information.