Skip to main content
Sindri is the Asgard platform's conversational AI entry point, letting internal users work through chat with the AI Agents built in Odin (Studio) to run queries, analysis, and system operations that require approval.

About Sindri | Asgard Agent Hub

Sindri is the Asgard platform's conversational AI entry point, the Agent Hub. It is built for ordinary users inside a company rather than the developers or administrators who build Agents. A user opens a conversation in Sindri, picks or delegates to an AI Agent the company has already put into production, and uses natural language to query data, analyse figures, and even carry out system writes that require human approval.

Sindri itself offers no interface for creating or configuring Agents. Administrators do that in Odin | Asgard Studio: designing an Agent's prompt, mounting the Semantic Models it can query and the MCP Servers and Skillsets it can call. Only once that is configured do those Agents appear in Sindri for users to talk to. The relationship is best understood as Studio being the back-office tool for building Agents, and Sindri the front-office chat interface for everyone else.

Core concepts​

The Sindri screen is built around a few concepts:

  • Workspace / Project: the workspace a user belongs to, holding several Projects. Each Project corresponds to one real business situation and one set of published Agents, so switching Project means switching business domain. See Project.
  • Available Agents: the Agents published in the current Project that you can talk to or delegate to directly. Each Agent card states what kind of request it handles. See Agent Hub home.
  • Directory: a folder that keeps several conversations and files on one topic together. See Directory.
  • My Chats: the list of conversations a user has opened. See My Chat.

Which Agents are actually available differs by Project. Go by what your company has published.