Indico Data releases industry-first large language model benchmark for document understanding tasks
Learn More
  Everest Group IDP
             PEAK Matrix® 2022  
Indico Named as Major Contender and Star Performer in Everest Group's PEAK Matrix® for Intelligent Document Processing (IDP)
Access the Report


Release Notes – Indico IPA v4.9

March 16, 2021 | Release Notes

Back to Blog

Thank you for being a valued Indico user! We’re constantly making updates to our app and APIs, working on new features, and garnering feedback to be best in class for intelligent process automation.  Have ideas on how to make our product even better – please let us know here!

Innovations and Updates in v4.9:
  •   Metrics for Review – Time on Task
    • In addition to the metrics that we released in v4.8 (which you can read about here), we have added a Time on Task and Time Saved metrics.
    • Time on Task captures the daily average time per document that Reviewers spend on documents, showing the difference between time spent on average in the Review Queue from what is spent on average in the Exceptions Queue. 
    • There are rolling 30 day metrics that also show overall the average time per document.


  •     Document Labeling Update
    • In 2020 we released document labeling, which allows you to keep the context and view of the document, and with simply applying highlights to your documents, you are training a model – and now we’ve made it even better and easier to use in both Teach and Review
    • We’ve rebuilt our canvas (and used even better technology) to make the mouse track easier across the screen and to allow the screens to load more quickly and accurately, especially for long or character dense documents.
    • Zoom on document is now available to see fine text or apply more refined highlighting.


    • The OCR text is available in Teach to easily validate that what is extracted is valuable and correct for the model being trained
    • Search now allows you to pick colors to have your search terms really pop on the page.
    • Our buttons for Next and Reject are now icons, and documents can now be skipped and be placed at the back of the randomized queue – a great alternative to reject if a Labeler is unsure about a document.


  •   Straight Through Processing for Review via Scripting
    • At the field level, we now support straight through processing per prediction based on criteria that match your use case.
    • Predictions can be accepted, rejected, or ‘No Value’ can be confirmed based upon variables chosen for the use case, and those scripted changes will be reflected in Review to the user, i.e. a prediction that is “auto-accepted” via scripting will appear as accepted in the Review interface, and similar behavior for rejected with the prediction not appearing in the interface. 
  •   Add Data to Workflow via API
    • Now that data can be added to a Dataset after creation, that additional data can be used to continue to train models by adding it to the Workflow via API. 
    • If you would like to use this functionality, please contact us and we will provide you information on how to do this.
  •   Configuration for Data Retention
    • Based upon your company’s data retention policy and storage options, we have built in configurations that can be mixed and matched to meet your needs.
    • Ability to configure your data retention policy for submissions with options based on time, if submissions are marked retrieved, if the results file has been accessed, and others.
    • If you would like a custom configuration for data retention and storage cleanup, please reach out to us.
  •   Document Bundles Support
    • Document bundles, or multi-file submission, allow you to submit multiple files under a single identifier and the files are retrievable under the single identifier.
    • Documents are searchable within the bundle.
    • Currently, there is support for Document Bundles in Workflows without Review.
  •   Other Updates and Bug Fixes
    • OCR settings can now be applied at the instance level.
    • Support for multiple SSO groups.
    • Prediction pills will no longer appear if predictions are turned off in Teach.
    • When a model is not yet trained or is currently training, the Class Balance chart will be the default view until Metrics are available after training in Explain.
    • For Extraction models, the span type will default to Overlap.
    • In Metrics for Review, the Average Age in Queue summary now appropriately represents the age of items when they were submitted.

Keep up to date on our latest happenings on our blog and our latest releases on our docs page!


Increase intake capacity. Drive top line revenue growth.


Get started with Indico

1-1 Demo



Gain insights from experts in automation, data, machine learning, and digital transformation.

Unstructured Unlocked

Enterprise leaders discuss how to unlock value from unstructured data.

YouTube Channel

Check out our YouTube channel to see clips from our podcast and more.
Subscribe to our blog

Get our best content on intelligent automation sent to your inbox weekly!