Subscribe

UiPath Document Understanding

UiPath Document Understanding

AI Center Relation to Document Understanding

Suggest Edits

AI Center is the infrastructure on top of which UiPath Document Understanding machine learning models run. These models can be deployed or instantiated for retraining with a few clicks (see this section).

See below a summary of the Document Understanding plus AI Center deployment options. Note that in this context "On premises" means the software is deployed in the customer’s environment, which can include Cloud deployments like Azure and AWS. “Cloud” refers to the UiPath-hosted Automation Cloud.

Deployment Type

Orchestrator

AI Center

DU Models

Licensing

Notes

SaaS

Any

Not used

Cloud (SaaS)

Cloud

  • Uses read-only SaaS endpoints (e.g., du.uipath.com/ie/invoices)
  • Models not retrainable
  • Community vs. Enterprise determined by API key
  • Enterprise version may store data temporarily for service quality assurance and monitoring, Community version retains data for UiPath use

Cloud

Cloud

Cloud

Cloud (in AIC)

Cloud

  • Full cloud deployment
  • Zero customer infrastructure footprint for AI Center/DU
  • DU models deployed via AI Center ML Skills. May store data temporarily for service quality assurance and monitoring

Hybrid

On premises

Cloud

Cloud (in AIC)

Cloud

  • Uses public ML Skill capability from AI Center
  • On-prem orchestrated Robots call public ML Skills hosted inside AI Center Cloud via direct URL, with DU API key from cloud platform
  • Reverse hybrid approach also possible (cloud orchestrator, on-prem AI Center) – same approach of exposing models from onpremAI Center using public ML Skills capability

On Prem - Connected

On premises

On premises

On premises (in AIC)

Cloud

  • Default licensing model when picking ”on-prem”
  • Cloud platform for licensing and metering checks only – no business data sent outside customer’s information security boundary (see this docs pagefor connectivity requirements)
  • Enables subsequent SKU swaps because of visibility on metering usage

On Premises - Airgapped

On premises

On premises

On premises (in AIC)

On Premises
(in Orchestrator)

  • Not the default licensing model for on-prem DU unless specifically requested –installation more complex than on-prem connected install
  • Licensing file created by DealHub with on-prem Orchestrator used for licensing
  • No post-deal close swaps as we have no visibility into actual usage

Using a Document Understanding model involves these steps:

  • Collect document samples and the requirements of the data points that need to be extracted.
  • Label documents using Data Manager.

📘

Note:

Data Manager itself will connect to an OCR Service.

  • Export labeled documents as a Training dataset to AI Center.
  • Export labeled documents as a Evaluation dataset to AI Center.
  • Run a Training Pipeline on AI Center.
  • Evaluate the model performance with an Evaluation Pipeline on AI Center.
  • Deploy the trained model as an ML Skill in AI Center.
  • Query the ML Skill from an RPA Workflow using the Document Understanding ML activity pack.

📘

Note:

Remember that using Document Understanding ML Packages requires that the machine on which AI Center is installed can access https://du-metering.uipath.com.

Updated 11 days ago

AI Center Relation to Document Understanding

Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.