Choosing the correct IXP capability

Frequently asked questions about choosing the correct IXP capability, covering the Unstructured and complex documents capability and when to use each capability.

Does the Unstructured and complex documents capability have the same functionality as Document Understanding Classic or Modern experience? How are they different?

The Unstructured and complex documents capability does not have any splitting or classification capabilities, it is extraction only. Additionally, it is a prompt-based capability that leverages generative AI and does not use any specialized pre-built models.

The capability does not learn from any user annotations or validation, which is the main difference compared to the Classic or Modern experience.

What does the Unstructured and complex documents capability mean for Document Understanding Generative Extraction activities?

Unstructured and complex documents supports documents up to 50 pages. Support for documents between 50 and 150 pages is in preview.

Meanwhile, if you need to process unstructured documents with more than 50 pages, you can use the Document Understanding Long Document Complex Layout generative extractor, available only in Studio. For more details on Document Understanding Generative Extraction activities, refer to Generative Extractor.

Note:

We will continue to support Document Understanding Generative Extraction activities, but we will no longer prioritize any further improvements to this capability.

What does the Unstructured and complex documents capability mean for AI Center?

AI Center is being deprecated in multiple phases:

• Phase 1 deprecation (Open-Source packages and out-of-the-box (OOTB) models) was announced in December 2024
• Phase 2 deprecation (Bring your own models) was announced in February 2025

It's important to note that UiPath will provide alternatives for each deprecated capability, including IXP, Integration Services, or UiPath GenAI activities. These new solutions are designed to replace the deprecated functionality with improved capabilities.

Additionally, UiPath will continue supporting all deprecated capabilities for some time, following the official announcement. Check the AI Center Upcoming deprecations page for additional information on specific timelines and suggested alternatives.

When should I use which capability?

The capability will vary based on the data and document types that need to be processed. Generally, we recommend using:

• The Communications Mining capability for email or ticket use cases.
• Document Understanding Modern or Classic projects for structured and semi-structured documents.
• The Unstructured and complex documents capability when:
  • You need to extract data from documents that contain paragraphs of free-form text or complex elements, that is, complex tables, graphics, charts, checkboxes, call-out boxes, signatures, handwriting, and more.
  • Your documents contain inferred values, which are values that are not extracted as-is from the document.
  • There is a significant variation in the layout of documents and fields that you need to extract.
  • You need to extract data from multiple document types, included as one in a stack of documents, without having to split them.

For further details on IXP capabilities and their intended use cases, refer to Choosing the correct capability.

Why wouldn’t I use the Unstructured and complex documents capability for any documents if it requires less training and time to deploy?

The Unstructured and complex documents capability can be used for any document type, including structured or semi-structured documents. However, the Unstructured and complex documents capability has been optimized for more complex use cases, where the traditional Document Understanding capabilities or extractors struggle.

Additionally, you might benefit from specialized OOTB models, custom model fine-tuning, and retraining for continuous improvement when using Document Understanding Modern or Classic projects.