Skip to main content

24 posts tagged with "KYRA Digital Agents"

View All Tags

KYRA:Docs: Reasons for Resolution Decisions and Extraction Rationale

KYRA:Docs now gives reviewers more control over the reasons recorded against resolution decisions, and shows the extraction rationale for every extraction. Reviewers can record an optional reason when they accept an extracted value or enter a custom value, configurators can choose whether a reason is mandatory when a reviewer discards, and the KYRA:Docs rationale is now shown for Match outcomes and for data groups after they have been actioned. No configuration is required to use the optional reasons and the rationale.

KYRA:Docs: Tenant-Level and Document Type Instructions

KYRA:Docs Managed Knowledge Base Instructions can now apply more broadly than a single Data Group or Field. Select Tenant (all documents) to write one Instruction that applies to every document in your tenant, or select Document Type to write one Instruction that applies to a specific Document Type, such as Passport, instead of repeating the same rule against each field. For example, a tenant-level Entity Data Extraction Instruction can tell KYRA:Docs to read all dates as DD/MM/YYYY, and a Passport Instruction can tell it to return the full four-digit year wherever the expiry date shows only two digits.

KYRA:Docs: Automatic Discard of Entities with a Discarded Parent

KYRA:Docs now automatically tags an extracted entity as Extraction automatically discarded — Parent extracted record discarded when all of its parent entities are discarded during review, and excludes it from the import. This replaces the previous behaviour, where a child of a discarded parent could be imported without a link to the hierarchy.

Tagged entities are read-only and need no reviewer action. Entities that still have at least one parent that is not discarded are imported as normal. No configuration is required.

KYRA:Docs: Execution Banner

KYRA:Docs now displays an execution banner at the top of the Document Data Extraction V2 task. The banner shows progress in real time while KYRA:Docs extracts information, including the step that is running and live counts of documents processed, data items and related parties extracted, and items that require review. When extraction finishes, it shows the final counts and a summary.

The task can now be opened before and during extraction, rather than only after extraction has finished. The banner is available to every user who has access to the task.

KYRA:Docs: Extraction of Fields with Trigger Conditions

KYRA:Docs now supports extraction of entity fields and data group fields that have trigger conditions. Administrators can turn on the new Include Trigger-Conditioned Fields switch on the Document Data Extraction V2 task, and reviewers see each trigger-conditioned field only when its trigger condition is currently met.

To use this capability, turn on Include Trigger-Conditioned Fields for the task in Journey Builder. The switch is off by default, so existing behaviour does not change until it is turned on.

KYRA:Screen: Materiality Assessment

KYRA:Screen now supports automated materiality assessment of confirmed screening hits on the Screening Resolution & Materiality task. Administrators configure Materiality Instructions to define how hits are assessed as Material or Immaterial for each screening category, and KYRA:Screen applies them automatically.

To use this capability, create and activate Materiality Instructions and enable the Complete materiality assessment toggle on the KYRA:Screen agent instance.

KYRA:Docs - Document Data Extraction V2 Now Available

KYRA:Docs — Document Data Extraction V2 brings the review of extracted entity data, data groups, and related parties into a single, unified task. Reviewers can now compare extracted information against the current entity record, understand the outcome assigned by KYRA:Docs, and record an explicit resolution before any changes are applied. This release replaces the legacy extraction experience with a more controlled, transparent, and auditable review process — supporting both standard and Maker-Checker journeys.

KYRA:Report

KYRA:Report is a new digital agent that lets users build and run reports using natural language. Rather than writing SQL, users describe the data they need in plain language and KYRA:Report generates the query, runs it, and returns the results ready to preview, refine, save, schedule or analyse. The agent is schema-aware, resolving the correct fields and parameters for your tenant, and retains conversation history so users can resume a session or reuse a query later.