Skip to main content
AutoSage is a platform for building grounded AI agents — assistants that answer from your documents and databases, use your connected apps and tools, remember what matters across conversations, and can run on a schedule or be called through an API. You can use AutoSage two ways, and this documentation is split to match:

Product docs

Configure knowledge bases, agents, chat, and memory from the dashboard. No code required.

Developer docs

Integrate AutoSage into your own product with the REST API, streaming, and API keys.

​
What you can build

Knowledge assistants

Upload documents and connect databases, then ask questions and get cited, grounded answers.

Autonomous agents

Give an agent a goal, tools, and a schedule, and let it run tasks on its own.

Customer-facing bots

Embed a chat experience for your end users, isolated per customer through a single API key.

Agent networks

Let agents delegate to other specialized agents to answer questions outside their own knowledge base.

​
How AutoSage fits together

Everything in AutoSage lives inside a simple hierarchy. Understanding it once makes the rest of the platform click:
1

Organization

Your company or team account. Holds billing, members, and everything below.
2

Environment

An isolation boundary — like dev, staging, and production. API keys belong to one environment.
3

Tenant

A resource owner with its own credits and quotas. In a customer-facing product, a tenant often maps to one of your customers.
4

Knowledge base

The container for a set of documents, database connections, prompts, and the agents that use them.

Learn the concepts

Start with the core concepts to see how knowledge bases, agents, chat, and memory relate.

Create an agent

Upload documents

Set up memory

Authenticate the API

Stream a chat

Invite your team