AI Assistant

Every Syntext doc site ships with an AI assistant trained exclusively on your documentation. It answers with citations, refuses to invent features, and reports what it couldn't answer as documentation gaps.

How It Works

After each production deploy:

Content is chunked by heading
Chunks are embedded (OpenAI embeddings) into a per-project vector namespace
Visitor questions retrieve the most relevant chunks by cosine similarity
The model answers grounded in those chunks, with source citations

Your content is isolated per project — it's never used to answer questions on other tenants' sites, and never used for model training.

On Your Doc Site

The assistant appears automatically on deployed sites — the search bar doubles as an "ask" box, and visitors can open a full chat panel. No configuration required.

While the assistant works, it shows its progress as thinking states ("Searching docs for…", "Found results for…") before streaming the answer, so visitors always know something is happening. Each conversation can be cleared with the trash button in the panel header, and every answer carries source citations plus an AI disclaimer.

Customization

Under Settings → AI Assistant in the dashboard:

Assistant name and persona/tone
Visitor rate limit — questions per visitor per window (details)

Embedding Anywhere

The chat widget can be embedded on any site — your marketing pages, your app, your support portal:

<script
  src="https://widget.syntext.dev/widget.js"
  data-project-id="prj_abc123"
  async
></script>

The widget fetches its config from the public config endpoint and streams answers from the chat endpoint.

In AI IDEs (MCP)

Connect your docs to Cursor, VS Code, Claude, and other MCP clients so their AI can answer from your docs. The fastest path is the hosted connector — one URL (https://api.syntext.dev/mcp), OAuth sign-in, no API keys — see MCP Integration.

Feedback & Gap Detection

Visitors rate answers 👍/👎 (feedback endpoint)
Low-confidence and unanswered questions are recorded as documentation gaps, ranked by frequency, in Analytics → AI Insights

Limits & Billing

AI queries are metered per organization per month. When the quota is reached, the assistant returns a friendly limit message and the API returns 403 USAGE_LIMIT_EXCEEDED — see Rate Limits.