The current Datafor conversational-analysis experience is a standalone AI Assistant opened from the AI Agent card on Home. It is no longer limited to a panel inside the report editor.

1. Open AI Assistant

1. Open Home.
2. Click AI Agent.

The page provides New Chat, History, Common Questions, Ops, and LLM in the top toolbar.

2. Select an analysis model

Before asking a question, choose the analysis model that contains the required metrics and dimensions. You can select a recently used model card or use Select Model in the composer.

After selection, Datafor displays a model brief, the number of available metrics and analysis dimensions, suggested questions, and a link to the complete metric list.

3. Ask an analytical question

Enter a question in the composer and send it. Make the request easier to execute by including:

• the metric;
• the time range;
• the grouping dimension;
• the required comparison, ranking, share, or trend.

Examples:

• Show monthly sales for last year.
• Rank product departments by margin, highest first.
• Compare units ordered with units shipped by warehouse class.

The answer is governed by the selected analysis model and the current user's data permissions. Check returned metrics, filters, and time basis before relying on the interpretation.

4. Continue or review conversations

• Click New Chat to start a separate conversation.
• Click History to reopen an earlier conversation.
• Use the model's suggested questions to begin with a request that matches the available metadata.
• Select a different model when the question belongs to another business domain.

5. Related administration

• Common Questions maintains reusable questions by analysis model.
• Ops shows usage and manages daily question quotas.
• LLM manages model profiles, runtime assignments, and schemes.

Troubleshooting

Symptom	Check
No model can be selected	Confirm the user can access at least one analysis model.
Model brief or suggestions do not load	Check AI Agent connectivity, LLM assignments, and the selected model's prepared metadata.
A query returns no data	Check the model, metric, time range, filters, source data, and row-level permissions.
The Agent reports a model-stage error	Open LLM and confirm that all 15 required assignments are complete and verified.
The progress title stays at Understanding your question until the whole answer appears at once	Datafor is running behind a reverse proxy that buffers the progress stream. Configure the proxy as described in Deploying Datafor Behind Nginx.