UiPath OnPrem Orchestrator with Document understanding implementation challenges

AI Engineer User Group

Hi All,

Implementing document understanding in on prem orchestrator UiPath set up using Automation suits . Facing various road blockers frequently in document understanding usecase.
Below are the listed challenges

  1. Getting dynamic behaviour in ml skills ,which runs for
    certain times and won’t work some other time.
  2. Pipelines will fails again and again based on increment in
    Dataset increase or ML skills increment.
  3. Using recommend infra set in our environments to avoid
    Infra issues.

Below are tried approaches to solve issue

  1. Tried to increase the infra requirements and won’t worked
    out.
  2. Increased CPU and GPU settings to achieve based on
    ML skills increment
  3. Used good quality of documents for OCR extractions
  4. Also utilized proper versions /compatible versions of
    studio,packages and orchestrator as per
    recommendations of UiPath.

Still issue is remains same as service delivery to business ,it’s a big challenge in document understanding implementation?

Need genuine answers and suggestions.
UiPath Community Kudos Orchestrator ai
ocr Academy Feedback Document Understanding uipath-action-center community

Can anybody help me on above challenges ?