Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/FloxTBoTyy/BoardPulse-AI/llms.txt

Use this file to discover all available pages before exploring further.

BoardPulse AI ships with an OpenAI-compatible API endpoint, which means any client that speaks the OpenAI chat completions protocol can connect to it — including Open WebUI. The default Docker Compose setup starts Open WebUI pre-configured to point at BoardPulse AI, so you can start asking questions through a polished chat interface without any extra setup.

​
Accessing Open WebUI

Open a browser and go to: 
http://localhost:3002
Open WebUI is available as soon as the Docker Compose stack is running. No login is required in the default configuration.

​
Selecting the model

In the Open WebUI model selector, choose boardpulse-executive. This model name is registered by the OpenAI-compatible endpoint and routes all requests through the BoardPulse AI pipeline.
1

Open the model selector

Click the model name at the top of the chat window to open the dropdown.
2

Select boardpulse-executive

Choose boardpulse-executive from the list. If it does not appear, confirm that the BoardPulse AI backend is running and that Open WebUI is pointed at the correct OPENAI_API_BASE_URL.
3

Ask your first question

Type a business question in plain language and press Enter. BoardPulse AI will respond with a formatted answer, data table, chart (if applicable), and export download links.

​
What the response looks like

The OpenAI-compatible endpoint formats the ChatResponse into rich markdown that Open WebUI renders inline:

Executive summary

A professional natural-language answer summarizing the query result.

Data table

Up to 20 rows displayed inline. A note appears when the full result set is larger.

Chart

If a chart was generated, it is embedded as a base64 PNG image rendered directly in the chat.

Export links

Clickable CSV and XLSX download links for the full result set.
A collapsible Technical Details (SQL) section at the bottom of every response shows the generated SQL, the executed SQL (with any applied limits), the active provider, and the name of the user who made the request.

​
Authentication

If BOARDPULSE_OPENAI_COMPAT_KEY is set, the endpoint requires a Bearer token. Open WebUI passes this token via the Authorization header automatically when configured: 
BOARDPULSE_OPENAI_COMPAT_KEY=your-secret-key
If BOARDPULSE_OPENAI_COMPAT_KEY is empty or unset, the endpoint accepts requests without authentication. This is fine for local development but should not be used in production.
Always set BOARDPULSE_OPENAI_COMPAT_KEY in production deployments. Leaving the endpoint open without authentication exposes your database query interface to anyone with network access.

​
Docker Compose configuration

The connection between Open WebUI and BoardPulse AI is configured via the OPENAI_API_BASE_URL environment variable in docker-compose.yml: 
openwebui:
  environment:
    OPENAI_API_BASE_URL: http://api:8000/openai/v1
    OPENAI_API_KEY: ${BOARDPULSE_OPENAI_COMPAT_KEY:-boardpulse-dev-key}
The api hostname resolves to the BoardPulse AI backend container within the Docker network.

​
User identity forwarding

Open WebUI forwards two headers with every request that BoardPulse AI uses for audit logging:
HeaderDescription
x-openwebui-user-nameDisplay name of the logged-in Open WebUI user
x-openwebui-user-emailEmail address of the logged-in Open WebUI user
These values appear in the Technical Details section of every response under Requested by, giving you a full audit trail of who asked what and when.
The 20-row display limit applies only to the inline table in the Open WebUI chat. Export files always contain the complete result set. Click the CSV or XLSX link in the response to download all rows.

Build docs developers (and LLMs) love

Get started for freeTalk to us