Quick Start
Create your first organization, provider, agent, and run it.
This walkthrough assumes the platform is already running — see Installation if you haven't set it up yet.
1. Register and sign in
Open the app and register the first account. The first registered user is provisioned as the platform administrator.
2. Create an organization and project
Every resource in UnderOcean belongs to a project, which belongs to an organization:
Organization → Projects → Agents / Flows / Knowledge Bases / IntegrationsFrom the dashboard, create an organization, then create a project inside it.
3. Connect an LLM provider
Go to Project → Settings → Providers and add a provider configuration:
- Choose a provider type (e.g. OpenAI, Anthropic, Ollama)
- Enter the model name (e.g.
gpt-4o,claude-sonnet-4-5) - Paste your API key — it is encrypted at rest and is never returned by the API afterward
- For a self-hosted or OpenAI-compatible endpoint, choose Custom and set the base URL
See Guides → Connect a Provider for more detail.
4. Create an agent
Go to Project → Agents → New Agent and configure:
- A system prompt
- The provider + model to use
- Temperature and max tokens
- Any tools the agent should have access to
5. Run the agent
Open the agent's chat view and send a message. Responses stream token-by-token in real time, and you can expand the inference steps panel to see the underlying thought → action → observation loop as it happens.
Next steps
- Build your first flow to orchestrate multiple agents and tools together
- Set up a knowledge base to give an agent Agentic-RAG retrieval over your own documents