CryspIQ update brings governed data to Teams & Copilot

CryspIQ has released version 3.3.0 of its data platform, adding access to governed enterprise data through Microsoft Teams and Microsoft 365 Copilot.

The update is aimed at organisations that want staff to ask questions of internal data and receive answers that can be traced back to source records. Responses in Teams and Copilot run under each user's Microsoft sign-in and are limited by Microsoft Entra ID security groups.

Copilot reads directly from CryspIQ, so users do not need to create and maintain a separate data source for Microsoft's assistant. Each answer also shows the underlying evidence.

Version 3.3.0 also focuses on data preparation and governance before information is made available for queries. The release is intended to shorten the path from a source file to a trusted data model while identifying data quality issues earlier.

Among the changes is a guided setup process in which Lumen, the company's AI assistant, proposes how a source file should map into CryspIQ's data model. Users can review those suggestions, preview the result, and save the mapping.

Another addition is earlier detection of data quality problems at the mapping stage. Records that fail checks are held for the relevant data steward together with the reason for failure.

The release also includes what CryspIQ describes as a self-healing parking lot. In practice, this means business records parked because related context records have not yet arrived can be loaded automatically once the missing records are available.

Governance controls

The update also strengthens governance and access controls. Lumen answers operate under each user's permissions, while an organisation's data, mappings, lookups, and settings remain separate from those of other customers.

A new Company Profile screen allows an organisation's logo to appear in the platform's navigation. The release also introduces delegated role management for data administrators, within limits set by each organisation.

Data quality reporting has also been revised. Every count on the Business Quality and Steward dashboards can now open the exact records behind it, while each issue is linked to its source and assigned data steward. Dashboard filters also remain in the link when it is shared.

The mapping page has been changed to let data stewards map fields directly, reducing reliance on administrators for some tasks. High-confidence mapping suggestions are applied automatically, leaving users to confirm only uncertain cases.

CryspIQ is targeting several groups with the release, including executives and managers who want direct answers in Teams and Copilot, data administrators responsible for source-to-target mappings, and cyber and IT teams managing data access and separation between organisations.

Dan Peacock, Co-Founder and Chief Executive Officer at CryspIQ, said the product update is intended to address concerns over the reliability of AI-generated answers when the underlying data is incomplete or inconsistent.

"Everyone wants to use AI on their own data, but the real question is whether you can trust what it tells you. That trust is earned in the data layer, long before anyone types a question. With v3.3.0, we've focused on getting those foundations right, so people can act on the answers with confidence," Peacock said.

The release is now live. CryspIQ's broader approach centres on using a single business-event data model to structure records from multiple systems so they can be queried consistently over time.

Vaughan Nothnagel, Co-Founder and Chief Technology Officer at CryspIQ, said: "What drove v3.3.0 was a simple idea: businesses should be able to get an answer they can trust, quickly, from the data they already have. That means getting new data in faster, catching problems before they cause trouble, and making sure that when someone asks a question, the answer shows its evidence and respects what that person is allowed to see. Businesses gain confidence, which is increasingly important and more challenging to realise. Teams spend less time chasing data issues and waiting on reports, and leaders can act on answers knowing where they came from."