Enterprise analytics company AtScale says it has been named an OpenAI Select Partner, extending its work on connecting AI applications with governed business data and consistent calculations of financial and operational measures.
The Boston-based company announced the designation on 20 October 2026, identifying its Universal Semantic Layer as the principal technology supporting its work with OpenAI products.
AtScale said the relationship would support organisations developing AI applications that need to interpret company information and calculate measures such as revenue, margin and customer retention using established definitions.
The appointment is reported on the basis of AtScale’s announcement. Independent confirmation of its individual Select Partner designation from an OpenAI-maintained partner directory has not been established, and the designation should not be interpreted as an endorsement of every AtScale product or performance claim.
According to AtScale, the OpenAI Partner Network provides a framework through which participating companies can build and deliver AI solutions for enterprise customers.
AtScale’s contribution focuses on the business information supplied to AI applications. Its Universal Semantic Layer allows organisations to define financial and operational measures centrally so that different software systems can use the same calculation rules.
That function addresses a practical problem in enterprise reporting. A business may use different definitions of revenue depending on whether it is measuring sales orders, invoiced amounts or income recognised under its accounting policies.
An AI assistant asked to build a report can produce an unsuitable result if it chooses a calculation without understanding which definition applies. The problem can arise even when the underlying database contains accurate records.
AtScale’s semantic layer is intended to provide established definitions for the measures an application requests. Its software also supports query optimisation and governed access to information held in enterprise data systems.
The company says it is connecting these capabilities with OpenAI products, including Codex, so that AI-assisted development and analytical workflows can use business measures already defined by the organisation.
One application involves asking a coding assistant to construct a dashboard. Without agreed definitions, the generated software might calculate a measure differently from an existing management report.
AtScale’s proposed approach is to make the same definitions available to the assistant and to other reporting applications. This is intended to reduce inconsistencies between dashboards generated by different tools.
The company describes work linking Codex to its Universal Semantic Layer. That work does not establish that every Codex user has access to the integration or that it has been deployed across a specified number of enterprise customers.
AtScale also introduced AtScale for AI on 15 October, five days before the partner announcement. The product suite includes capabilities for interpreting business questions, applying calculation rules and assessing the accuracy of AI-generated analysis.
Within that offering, AtScale Ava is designed to clarify ambiguous questions. A request for revenue figures, for example, may require additional information about the reporting period or the accounting measure concerned.
AtScale’s AI Computation Engine, known as ACE, applies business definitions and calculation logic when generating requested measures.
The two functions address different stages of analysis: establishing what the user wants to measure and performing the calculation according to the organisation’s rules.
AtScale says its approach can be used with existing business intelligence applications as well as AI assistants. That allows organisations to maintain a consistent set of definitions across different reporting environments.
The effectiveness of this arrangement depends on the accuracy and maintenance of the underlying business definitions. A semantic layer cannot independently establish whether the organisation’s accounting policies, source records or metric definitions are correct.
AtScale chief executive Chris Lynch said organisations adopting AI for routine business decisions needed to avoid having different agents generate conflicting interpretations of measures such as revenue, margin and customer churn.
Company president Jay Schuren identified consistency between AI-generated dashboards and established financial reporting as an objective of the technology.
AtScale has separately published its own benchmarking research on the effect of semantic layers on AI-generated analytical answers. Those results are company-reported tests and do not establish independently verified accuracy across every model, customer environment or business question.
The October partner announcement does not include independently audited customer outcomes or a verified performance improvement attributable to the new relationship.
AtScale intends to expand its OpenAI-related offerings and pursue commercial activities aimed at helping shared enterprise customers evaluate and implement AI systems.
It also plans additional development involving AI-assisted semantic modelling and accuracy evaluation. These are future activities described by the company, rather than evidence that the proposed capabilities have been delivered to every customer.
The financial terms of the relationship have not been disclosed. AtScale has not identified a new contract value, a confirmed number of customers adopting the integrations or a revenue forecast arising from the appointment.
The confirmed development is AtScale’s announcement of its Select Partner designation and its plans to extend governed business data capabilities into OpenAI-related workflows. Wider adoption and improvements in analytical performance remain to be demonstrated through customer deployments and independently assessed results.
You must be logged in to post a comment.