> ## Documentation Index
> Fetch the complete documentation index at: https://docs.realtimelca.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Assistant

> A conversational AI built natively into Real-Time LCA, powered by the MCP server.

The **AI Assistant** (also called the **SI Agent**) is a conversational assistant built natively into Real-Time LCA and powered by the [MCP server](/user-guide/integrations/ai-tools/mcp-server). Ask questions about your projects and trigger workflows directly in the platform, without switching to an external tool.

The AI Assistant is an **open beta, enabled per organisation**. It is off until an administrator requests access.

## What you can do

* Ask about a project's emissions, mappings, or dashboard in plain language.
* Trigger in-platform workflows — such as mapping materials or generating a report — from a conversation.
* Stay in one place: because it runs inside Real-Time LCA, you don't need to connect an external AI client.

## Models and data

The assistant uses **Mistral** models through the Mistral API, processed on EU servers, and your data is never used to train them. Everything you put in a prompt — plus the project data the assistant decides to query — is sent to the provider, and Real-Time LCA stores prompts for up to 30 days. See [Transparency & privacy](/user-guide/integrations/ai-tools/transparency-and-privacy) for the full picture. As with every AI feature, AI-generated output is labelled and should be reviewed before you rely on it.

<Warning>
  Never put personal information into a prompt — your prompts, and the project data the assistant queries to answer them, are sent to the AI provider.
</Warning>

## Get access

<Card title="Request access to the AI Assistant" icon="envelope" href="/get-involved/beta-programs/si-agent" horizontal>
  The AI Assistant (SI Agent) is an open beta — enabled per organisation on request.
</Card>
