AI Chatbot

Let your end-users ask questions of their data in natural language — and get back both text explanations and live charts, without ever leaving your product.

The AI Chatbot is a ready-made chat experience you drop onto the same page as your existing Embeddable dashboard. End users type a question; the chatbot queries their data, renders charts, and can even add, edit, resize, or reorder charts on their Custom Canvas on their behalf.

How it works

The AI Chatbot is made up of three pieces that work together:

• <em-ai-chat-bot> — a web component you drop onto your page alongside <em-beddable>. It connects to the AI orchestrator via WebSocket and authenticates using the same embeddable tokens you already generate for your users.
• AI Orchestrator — hosted by Embeddable. Receives messages from the chatbot, calls your LLM using the API key you've saved in the platform, queries your data models, and streams responses back.
• Platform LLM config — a single API call (PUT /api/v1/llm-config) where you save your LLM provider API key. This is how you bring your own key (BYOK).

End User types a question
  │
  ▼
<em-ai-chat-bot>  (WebSocket)
  │
  ▼
AI Orchestrator  (Embeddable-hosted)
  │  ├── fetches LLM API key from platform (encrypted at rest)
  │  ├── queries your data models via /ai/v1/* endpoints
  │  └── enforces row-level security from the embeddable token
  │
  ▼
Response streamed back → text + rendered chart

Prerequisites

Before adding the AI Chatbot you need:

1. Custom Canvas enabled on your dashboard — the chatbot uses Custom Canvas datasets and templates to render charts.
2. At least one dataset enabled in Custom Canvas and at least one chart template configured. The more templates you define (and the more open their inputs), the more chart types the AI can offer.
3. A security token generated server-side for each end user — see the Tokens API. The chatbot uses this token exclusively for authentication; no separate chatbot credential exists.

Setup

curl -X POST https://api.eu.embeddable.com/api/v1/llm-config \
  -H "Authorization: Bearer <your-workspace-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "anthropic"
    "model": "claude-sonnett-4-0",
    "apiKey": "<your-llm-api-key>"
  }'

<em-beddable
  id="my-dashboard"
  token="<security-token>"
  embeddable-id="<embeddable-id>"
/>
 
<em-ai-chat-bot
  embeddable-tokens="<security-token>"
/>

The end-user experience

End users interact with the chatbot in plain language. The AI can:

• Answer data questions — retrieve and summarise data from any dataset available in Custom Canvas, presented as text or charts.
• Add charts to the canvas — place new charts based on what the user asks for.
• Edit existing charts — update a chart's dataset, measure, or filters on request.
• Resize and reorder charts — rearrange the canvas layout conversationally.
• Delete charts — remove charts the user no longer needs.

All responses are scoped to the end user's security context. The orchestrator rejects connections where tokens belong to different workspaces, so there is no risk of cross-user or cross-workspace data leakage.

Security

Related

• Custom Canvas — required prerequisite; also where chatbot-created charts appear
• AI Endpoints — power your own chat UI instead of using the ready-made component
• Tokens API — how to generate the security tokens the chatbot relies on
• Row-level security — how per-user data access is enforced