Arango logo
arango.ai 

Set up AutoGraph

Set up AutoGraph using the web interface or the HTTP API to build knowledge graphs from your enterprise documents

AutoGraph is available in the Arango Contextual Data Platform through two interfaces:

  • Web interface: A guided, step-by-step workflow for creating projects, uploading documents, and deploying retrieval services without writing code.
  • HTTP API: Full programmatic control over the AutoGraph pipeline for automation and integration into existing workflows.
Both interfaces produce the same result: a fully operational knowledge graph with deployed retrieval services. Choose the web interface for a guided experience, or the API for automation.

Supported file formats

The corpus build does not parse documents itself. It hands every input that is not already plain text or Markdown to the File Parser service, which samples it and converts it to Markdown. AutoGraph then embeds and clusters that Markdown.

The File Parser is an internal data platform service installed once per environment, without a web interface of its own. The Importer uses the same service when it builds the knowledge graph, so the whole AutoGraph pipeline accepts one consistent set of inputs:

  • PDF: .pdf files including scanned documents, read using OCR
  • Office documents: .doc, .docx, .ppt, and .pptx files
  • Plain text: .txt and .md files

For the per-format detail, see Format support.

Parsing runs on CPU. Throughput for large-scale ingestion of PDF and Office documents is governed by how many File Parser worker pods your node pool can support, not by GPU availability. Parsing PDFs is by far the most expensive operation, because every page without a text layer has to be read using OCR. See Tuning the File Parser if a self-hosted cluster is slower than you expect.
A document has to yield extractable text. An image-only PDF without OCR-readable content fails as a single file, and the build can still complete with error_code: FILE_PARSER_PARTIAL_FAILURE, see Document parsing.

Prerequisites

  • Arango Contextual Data Platform
  • LLM and embedding API access (commonly OpenAI-compatible endpoints)
  • Valid JWT for the API (Authorization: Bearer ...)
  • Platform auth reachable from the service (for token validation and renewal), if your deployment uses it

Before importing data, you need to create a project. Projects keep datasets and configurations isolated from each other.

For detailed instructions on creating and managing projects, see the Projects section in the Arango Control Plane (ACP) documentation.

To install and start the AutoGraph service, use the /_platform/acp/v1/autograph endpoint of the Arango Control Plane (ACP).

Get started

The web interface of the Arango Contextual Data Platform lets you create, configure, and run a complete AutoGraph workflow through a streamlined web interface.

  1. Navigate to Agentic AI Suite > AutoGraph Studio in the sidebar.
  2. Create a new project and upload your documents into categories.
  3. Configure your LLM provider, embedding, and multimodal models.
  4. Start the build to deploy the AutoGraph service and build the Corpus Graph.
  5. Generate strategies and deploy an AutoRAG retriever to query your Context Graph.

For the full walkthrough, see the Web Interface guide.

The AutoGraph service exposes HTTP endpoints (port 8080) for programmatic access. The recommended call sequence is:

  1. Upload files to the File Manager under the scope [project, category]

    POST https://<EXTERNAL_ENDPOINT>:8529/_platform/filemanager/_db/:database/rag-input

  2. Build corpus with those category labels in categories

    POST https://<EXTERNAL_ENDPOINT>:8529/autograph/v1/corpus/builds

  3. Generate strategies

    POST https://<EXTERNAL_ENDPOINT>:8529/autograph/v1/rag-strategizer/analyze

  4. Orchestrate import

    POST https://<EXTERNAL_ENDPOINT>:8529/autograph/v1/orchestrate

Steps 2 to 4 are asynchronous and acknowledge the request with 202. Poll the matching status endpoint before you move on to the next step.

Authentication uses JWT Bearer tokens. For full endpoint documentation, see the API Reference.

Learn more

  • Use Cases: Real-world enterprise applications and business impact metrics.
  • Architecture: Learn more about the three-layer knowledge graph architecture and resulting collections.
  • Design Guide: How to structure your data with categories, layers, and components.

On this page