Assistants

This page covers creating an assistant that can chat with your dataset and connecting a language model to power it. If you haven’t created a project and set up your search yet, start with the Importing Data guide and set up the semantic search first.

Create an assistant

If you follow our Quick Start process, you will create an assistant alongside your project (unless you choose to skip the assistant step). TellusR automatically configures this assistant to answer questions grounded in your project’s documents.

Start page

The Quick Start process always creates a Chat Assistant. TellusR also supports other assistant types - see the Assistants-page documentation for the full list and instructions on how to create them.

Assistants Manager

Choose and connect a language model

Your assistant needs a language model behind it before it can generate answers. TellusR can utilize a wide range of providers, including OpenAI’s GPT, Anthropic’s Claude, and Mistral, or your own models, hosted locally or in the cloud. Pick whichever fits your needs, then connect it, usually just by supplying the API token you received from that provider.

See the LLM Config documentation for the full list of supported providers and how to connect to them.

Start page