Reasoner Quick Start

Turn slow AQL queries into fast ones - let AI find the bottleneck and rewrite the query, with results verified to match

Beta

Prerequisites

An ArangoDB instance with the ArangoDB MCP Server running and connected.
An OpenAI API key.
An Agentic AI Suite license (for the web interface).

Optimize your first query

Open the Reasoner

In the Query Editor, click Open Reasoner in the Welcome tab - or click Optimize from any query tab.

Start the service

In the Setup Reasoner panel, enter your OpenAI API key, select a model, and click Start Service. Wait for the MCP Connected status.

Describe what you want

Type your request, for example:

Optimize this query: FOR u IN users FILTER u.status == 'active' SORT u.created_at DESC RETURN u

Press Enter.

Review and apply

Read the AI Reasoning Response with the optimization report, then click Open in Editor to apply the rewritten query.

With the Reasoner service running (port 8080, MCP connected), send your request:

Send an optimization request

POST https://<EXTERNAL_ENDPOINT>:8080/query

{
  "request": "Optimize this query: FOR u IN users FILTER u.status == 'active' SORT u.created_at DESC RETURN u",
  "stream": true
}

Consume the response

With "stream": true, consume the Server-Sent Events (phase, tool_call, tool_result, text, done) to follow progress. Set "stream": false to receive a single JSON response whose result field holds the full markdown report.

You now have an AI-generated optimization report: performance metrics, a rewritten query, and validation that it returns identical results.

Next steps

Web Interface: The full guided experience.
API Reference: Endpoints, streaming, and event types.