Users can filter the audit log list by date range, activity name, operator (human or machine), and username. 

Users can download a CSV file of their audit logs with the exact same information as the UI.

Nothing about the existing audit logging UI has changed. Users will simply see more events logged and more Activity Types in the drop down menu.

Security

Data Integrity

Scalability

Reliability

Native ML models for understanding handwriting across 15 languages

Ground-breaking point & click simplicity in training, supervision & reporting for business users

Scalable AI platform to process billions of documents across any industry

Ready to train models for unique, unseen enterprise data or sensitive, secure forms of data

Supports all deployment options: 
On-premises, hybrid and SaaS

Adheres to the highest standards of FedRAMP, WCAG 3.0 for automation in secure, compliant & private world

Accelerate TTV with co-pilot features that assist in training, QA & supervision

Execute hyperautomation workflows via hundreds of predefined & custom integrations

Orchestrate accuracy and automation levels with seamless human in the loop 

Scenario:  Auditors often review freeform transaction reports that document multiple occurrences of the same transaction type across different dates. For instance, a report might state, "On March 3rd, a $10,000 wire transfer was sent to account X. On March 15th, a similar transfer was made to account Y."

Long Form Extraction can automatically capture all instances of wire transfers and corresponding dates, providing a comprehensive view for auditors without requiring manual data entry.

Banking | Financial Audit Reports

Scenario:  Government agencies often manage and review numerous grant applications that contain repeated references to funding requirements, project timelines, and beneficiary details. For instance, a grant application might state, "The project will commence on July 1st, with quarterly reports due on October 1st, January 1st, and April 1st.

Long Form Extraction can automatically extract these key dates and funding amounts mentioned multiple times throughout the application, simplifying the review process 
for government officials and ensuring that all necessary details are accurately captured for decision-making and compliance monitoring.

Public Sector | Grant Applications & Funding Review

Scenario: Insurance companies often review policyholder communications during claims processing, where multiple interactions occur over time. For example, "The policyholder first reported the incident on August 5th, followed up on August 10th, and submitted the final documents on August 15th."

Long Form Extraction can capture each date and interaction, ensuring that all touchpoints are documented and reducing the likelihood of missing critical 
information in the claims review process.

Insurance | Policyholder Interactions

Scenario:  Law firms dealing with joint ventures or partnership agreements often encounter legal documents where multiple parties are listed throughout the text. 
For example, a document might state, "Party A agrees to provide service on April 1st, and Party B will deliver products by May 1st." 

Long Form Extraction can identify and extract all occurrences of the parties involved 
and the associated deadlines, ensuring that every relevant detail is captured for 
contract management and review.

Corporate Law | Joint Legal Agreements

Field Level
Accuracy Targets

Set target accuracy rates at the field level on structured forms vs the entire document to focus resources on the most important data.

Layout ID Matching
For Structured Forms

Ensure a consistent layout match based on the unique identifier on your forms, increasing automation and accuracy at the same time

Extended Model 
Compatibility

Preserves investments in previously trained models by enabling seamless model portability across different versions of the Hypercell platform

Document Drift 
Management

Automatically identifies and updates models to handle new layouts without requiring human intervention.