- Overview
- Model building
- Model evaluation
- Model deployment
- Consuming models
- Developer
- Frequently asked questions
Runtime
Frequently asked questions about runtime in IXP in Flow, covering model consumption, confidence scores, and validation.
Are the models of Extract nodes consumed via robots or standalone Document Understanding APIs?
In Maestro Flow, the Extract node runs the model as part of your flow. Outside of Flow, you can also consume the model via robots and the standalone Document Understanding Automation Cloud API. For details, check Consuming models via automations and Consuming models via Document Understanding API.
How can I consume the outputs of the models of Extract nodes?
In a flow, downstream nodes consume the output of the Extract node. For details, check Managing Extract nodes.
Outside of Flow, model outputs can be consumed via APIs, Document Understanding Activities, or IntelligentOCR (IOCR) Activities. Make sure to:
- Install the IOCR package v6.23.2-preview or newer and select the Document Understanding Project Extractor activity
- Install the Document Understanding package v2.13.2-preview or newer and select the Extract Document Data activity
Does IXP in Flow provide confidence scores? How are they calculated?
Yes, we provide confidence scores in the model response and in Action Center (runtime), but not in the IXP user interface (design-time). We use the token probabilities returned by the LLM to estimate field and extraction level confidences.
Models that do not return log probabilities always return a confidence value of 1.0 for all predicted values. For the list of affected models, check the Choosing the most suitable model section.
Can the outputs of the models of Extract nodes be consumed from hybrid deployments, that are on-prem Orchestrator or Automation Suite and Cloud IXP model?
Yes, this is supported via External Apps. For more details, check OAuth apps.
Can the outputs of the models of Extract nodes be consumed from different organizations or tenants, that are Production, Development, or User Acceptance Testing (UAT)?
Yes, this is supported via External Apps. For more details, check OAuth apps.
Does the validation station work with IXP in Flow?
Yes, it works the same way as Document Understanding Modern or Classic projects.
The Validate extraction node works only with Extract nodes created in the same flow, with the Extract document option. It does not yet work with Extract nodes bound to a published model or to a model from another flow.
Do the models of Extract nodes learn based on human-in-the-loop (HITL) validation from Action Center?
No, there is no feedback loop between Action Center and the models of Extract nodes. This is planned to be supported in future releases alongside fine-tunable models.
- Are the models of Extract nodes consumed via robots or standalone Document Understanding APIs?
- How can I consume the outputs of the models of Extract nodes?
- Does IXP in Flow provide confidence scores? How are they calculated?
- Can the outputs of the models of Extract nodes be consumed from hybrid deployments, that are on-prem Orchestrator or Automation Suite and Cloud IXP model?
- Can the outputs of the models of Extract nodes be consumed from different organizations or tenants, that are Production, Development, or User Acceptance Testing (UAT)?
- Does the validation station work with IXP in Flow?
- Do the models of Extract nodes learn based on human-in-the-loop (HITL) validation from Action Center?