Hosted
Start with calypso-rag-agent
Use the default hosted RAG agent as a production answer layer. It resolves its saved policy at runtime and searches the buckets you attach.
Calypso Agents
Hosted RAG agents with curated memory.
Calypso Agents search selected Buckets, answer with citations, and deploy through the same OpenAI-compatible interface your product, workflow, or website already knows how to call.
hosted
RAG runtime
scoped
bucket memory
cited
answers
Agent profile
calypso-rag-agent:support
memory
support-handbookrelease-notes
sample answer
“Install the widget script, pass your team id, then verify the source references in the response.”
3 grounded sources
API
compatible
MCP
ready
Profiled
Create named agents for each job
Spin up profile-style agents such as support, pricing, onboarding, or legal with their own retrieval scope and answer behavior.
Grounded
Answer with citations and source context
Agents retrieve from bucket-backed knowledge and return source-backed answers users can inspect instead of trusting a black-box response.
Reusable
Deploy the same agent everywhere
Use one configured agent across your website widget, playground, OpenAI-compatible API, MCP, n8n workflows, and product UI.
Agent profiles
One hosted runtime, many specialized agents.
Give support, pricing, onboarding, and internal teams their own named agent profiles. Each profile can search a curated bucket set and answer in the voice of that workflow.
support
calypso-rag-agent:supportsupport-handbookrelease-notes
Answer customer troubleshooting questions with citations.
pricing
calypso-rag-agent:pricingpricingprocurement-faq
Explain plans, limits, and packaging from approved sources.
onboarding
calypso-rag-agent:onboardingcustomer-onboardingimplementation-guides
Guide new users through setup steps and implementation details.
API workflow
Call agents like a model, with memory already attached.
Responses API
# Ask a named Calypso RAG agent
curl https://api.calypso.so/v1/responses \
-H "Authorization: Bearer sk_..." \
-H "Content-Type: application/json" \
-d '{
"model": "calypso-rag-agent:support",
"input": "How do I connect the widget?",
"store": false
}'Choose buckets
01Attach the source collections this agent should search.
Set the policy
02Control retrieval, citation visibility, and answer behavior.
Ask anywhere
03Call the agent from chat, API, MCP, n8n, or your own UI.
Show evidence
04Return grounded answers, sources, and retrieval metadata.
Deploy everywhere
One agent configuration, every answer surface.
Configure the agent once, then reuse it wherever your users or automations ask questions. The same bucket memory and citation behavior follows the model id.
Website widget
OpenAI-compatible API
MCP clients
n8n workflows
Product UI
Internal tools
Get started with Agents
Ship source-grounded AI agents without rebuilding RAG.
Create buckets, attach them to named agents, and launch grounded answers across your website, API, MCP clients, n8n workflows, and product experiences.