The Modern Data Company Launches Dex, Conversational AI That Knows When It Has the Answer and Admits When It Doesn’t

The Modern Data Company introduced Dex, a conversational AI experience built for DataOS®, the company’s enterprise data platform that works across existing infrastructure. Dex lets business users ask questions of their enterprise data in everyday language and receive explainable answers grounded in governed data products. Each answer draws on established business definitions, context and underlying sources. When the available data cannot support a reliable answer, Dex makes that limitation clear rather than generating an unsupported response. Dex is available now to DataOS customers.

Also Read: AiThority Interview with Tim Jurka, VP Engineering at LinkedIn

Enterprises have invested heavily in data platforms, yet getting a reliable answer to a business question can still require considerable effort. Connecting a general-purpose AI model to enterprise data makes that data more accessible, but access alone doesn’t provide the business understanding needed to interpret it correctly. Models need to know how metrics are defined, how data relates across systems, and which information can be trusted. Without that context, even a convincing answer can be inaccurate or misleading.

Interim findings from The Modern Data Company’s third annual Modern Data Survey highlight the challenge. Among more than 500 respondents, 61% said reliable business context is necessary for AI agents, yet only 16% reported deliberately engineering that context into their data.

Answers Grounded in Business Context

Dex, short for Decision Exchange, draws on the business context, data quality and governance built into DataOS data products. These reusable, governed products carry business definitions, metrics, relationships, ownership, quality controls and access policies, giving people and AI a consistent understanding of enterprise data. An ontology captures the business meaning and relationships behind the data, allowing Dex to interpret questions in the context of how an organization defines and operates its business.

Dex provides visibility into the metrics, definitions, filters, data products and underlying sources used to produce its answers. Because business definitions are established in governed data products rather than recreated in individual prompts, organizations can maintain greater consistency across users, teams and applications. When the available governed data cannot support a question, Dex identifies the limitation instead of presenting an unsupported answer as fact.

“Every enterprise wants AI that can give real answers about its business, but connecting a model to data isn’t the same as giving it understanding. A model needs to know how the company defines its metrics, which data sources it trusts, and how its data fits together. Without that, it guesses, and a confident guess is hard to catch,” said Saurabh Gupta, president and CEO of The Modern Data Company. “Dex gets that understanding from DataOS, where business definitions, trusted sources and data relationships are already built into governed data products. People can ask a question in plain language and get an answer they can trace back to the definitions and data behind it. Answers people can verify are answers the business can act on.”

Governance Built Into Every Query

Dex inherits the governance and access controls established in DataOS. Rather than relying on instructions to a language model to determine what information a user should see, DataOS enforces access policies as queries execute. Identity and permissions are evaluated at runtime, ensuring users can access only the data they are authorized to use.

Dex also supports natural follow-up questions, allowing users to explore an issue without starting over or losing the context of the original request. A sales leader investigating a decline in conversion, for example, can ask which regions were most affected, whether product availability contributed to the decline, and then request a visualization of the results. Each follow-up draws on the same governed data foundation, with the relevant definitions and access controls maintained throughout the conversation.

Works Across Existing Data and AI Infrastructure

Dex is designed for organizations whose enterprise data spans multiple platforms. Built for DataOS, it can work with data in Snowflake, Databricks, BigQuery, Iceberg, operational databases and APIs without requiring organizations to migrate or consolidate that data into another system. Organizations can also choose from supported commercial, enterprise and private LLMs based on their security, privacy and performance requirements.

Also Read: Architecture, Not Intelligence, Will Define the Future of AI Headphones

[To share your insights with us, please write to psen@itechseries.com]