Using IQ Bot
IQ Bot provides cognitive (intelligent) automation to uncover and transform important, but less structured data to automate business processes quickly and efficiently, simultaneously reducing human error.
Cognitive automation processes semi-structured and unstructured data and converts it into structured data that is used by Robotic Process Automation (RPA) bots for end-to-end automation.
Start using IQ Bot by creating a learning instance, which defines the type of document you must process, the language of documents, and a list of data fields to capture and extract from each document. Next, train the documents and review the results of the training. You can also download the extracted data to a CSV file for review. After correcting any errors, save the botand send it for production. In the production environment, run the trained bot against a set of documents to automate data extraction.
Phases in IQ Bot business process
- Preprocess documents.
- Receive text segmentation and optical character recognition (OCR).
- Classify documents in groups.
- Extract document data.
- Validate and correct failed documents.
- Complete validation and save.
- Trigger approval.
- Obtain final review and approval.
Users who create and configure automation tasks and deploy TaskBots also create IQ Bot learning instances, deploy the learning instances from staging to production environments, and correct documents with exceptions.
- Services users: Automation experts who train bots on sample documents, so these bots can later automatically process a larger volume of documents.
- Validators: Use a visual interface to manually verify or fix data extraction from a document.
- Common administrative tasks of Automation Anywhere Control Room
- Differences between structured, semi-structured, and unstructured documents
- Standard fields in a semi-structured or unstructured document, for example, invoice number, invoice date, and so on.
- General automation commands in Automation Anywhere
- Internet information services
- How to start and stop web services
- How to block and unblock ports
Use the general process for IQ Bot to create a learning instance, upload documents, build/train the bots, validate extracted data and make corrections, and set the bots to production.
- Create a learning instance and upload sample documents.
- After the documents are analyzed, review the report in the Performance report page. The report shows you important insights about your sample documents, for example, similar documents that can be grouped together, document groups that return all required fields, and document groups used to create and train learning instances.
- After the sample documents are analyzed, Train a learning instance by mapping required fields and setting validation rules for a document in a group that best reflects the documents in that group. When the learning instance is deployed in production, it processes all documents identified as part of this group.
- After training, Set learning instance to Production, and then use the botcommand to upload documents to the production environment for processing. See Upload documents to a learning instance.
- Any documents that do not complete straight-through processing because of field extraction or rule-related problems require human validation. Users are required to Validate document with errors.
- Throughout the process, use the IQ Bot Dashboard to monitor the progress of the production instances.
Throughout the process, use the IQ Bot Dashboard to monitor the progress of the production instances.
Features and benefits of using IQ Bot
Use IQ Bot for the following features:
- Train the learning instances in order of importance on the web-based Designer.
- Reduce setup time for new use cases with the Domain Management utility.
- If you are logged in to the Control Room with the Single Sign-on feature, you can open the IQ Bot Portal directly.
- Leverage the stronger security features in the Automation Anywhere platform as part of IQ Bot Version 6.0 integration.
- Exception handling is fast and seamless with the web-based Validator.
- Preview the data extraction results to verify the training provided to the learning instance.
- Use semantic analysis and automated classification to analyze and extract
data types and formats from learning algorithms, invoices, purchase orders,
and bills. It also does the
- Autodetects file values after field mapping
- Autocorrects exceptions from human expertise
- Flags exceptions based on the built-in confidence levels mechanism
- Leverage OCR technology, document classification, and data extraction of documents.
- Intelligent Document Processing with IQ Bot
- Unleash Your Intelligent Bots: Develop Cognitive Bots Driven by AI & Machine Learning (IQ Bot)