Aiding AI development

With enterprises often struggling to develop AI tools reliable enough to put into production, many data management and analytics vendors are making it a priority to provide customers with capabilities that better enable them to build trustworthy applications.

CData is moving from being the company that connects enterprise data to becoming the governed runtime data layer that AI developers and agents can consume directly.

In particular, connecting agents with the data and business logic that gives them their contextual awareness has been an emphasis, with AWS, Databricks, Microsoft and Snowflake all addressing context this month alone.

CData addressed context when it launched new Connect AI capabilities in March. Now, the vendor is trying to help customers more successfully build AI tools by simplifying the developer experience.

CData commissioned a study of more than 200 data and AI leaders in late 2025, and the results of that survey helped provide the impetus for building a Developer Edition of Connect AI, Connect AI Python SDK and CData CLI, according to Amit Naik, the vendor's vice president of AI architecture.

"Our customers told us [what they wanted], and then our own research confirmed it," he said, noting that nearly three-quarters of the survey respondents reported spending significant time on data integration rather than training models and building applications.

"Developers were asking for better ways to get started without going through an enterprise procurement cycle," Naik continued. "The Developer Edition, the Python SDK and the CLI are our direct response to that signal."

Connect AI reduces the need to integrate data by exposing enterprise APIs as a data layer with a standardized schema and read/write support to retrieve and update data. In addition, it includes support for MCP servers to standardize connections between AI models and proprietary data sources and Toolkits, a feature that enables developers to package governed data access into a single MCP server URL so agents only call on the data they need.

Connect AI Python SDK further simplifies access to data by enabling developers to import governed data into Python workflows without having to change the code, and CData CLI simplifies managing access to an enterprise's myriad data sources by providing an interface for CData's network of connectors.

"Connect AI Developer Edition stands out as the most valuable because it's free, includes the full enterprise feature set, and works immediately with popular AI coding assistants like Claude Code, Cursor, and LangChain," Catanzano said. "The Toolkits feature is particularly powerful."

Ni likewise highlighted the Connect AI Developer Edition.

"While the Python SDK and command line interface make developers more productive, more importantly, the Developer Edition promises an enterprise data layer that goes beyond developers to support every AI application and agent, enabling consistent development," he said.

Measuring up

With the launch of Connect AI last year and its updates since, CData is carving out a niche for itself, according to Ni.

Vendors such as Databricks and Snowflake -- along with Qlik and Teradata, among others -- are building broad platforms designed to handle complete data management and AI development lifecycles. CData, instead, is focusing more narrowly, attempting to distinguish itself with a layer that easily and efficiently connects AI with data.

"Every enterprise AI architecture needs a trusted way to connect AI intelligence with operational systems," Ni said. "That's the market CData has been chasing with one of the largest libraries of enterprise connectors. They aren't trying to outbuild Databricks or Snowflake. They’re trying to own one of the architectural control points those platforms still depend on."

Another way CData is differentiating itself is by combining a developer-first approach with enterprise-grade governance, according to Catanzano.

"This release … positions CData at the intersection of data infrastructure and AI development rather than just traditional integration," he said.

Looking ahead, CData's roadmap is focused on adding governance capabilities so AI agents can be trusted to act within organizational guidelines, according to Naik.

"As AI agents become more autonomous, the controls around data access have to keep pace," he said. "We're investing in the audit trails, access policies, encryption and compliance capabilities that enterprise teams need before they'll put an agent anywhere near production data."

In addition, product development initiatives include improving the quality of semantic models to enhance the contextual awareness of agents and advancing platform maturity to broaden the CData ecosystem, Naik continued.

Catanzano, meanwhile, suggested that CData could continue serving existing customers and perhaps appeal to new ones by building pretrained semantic layers or domain-specific data models for common enterprise tasks such as financial reporting or customer experience.

"Additionally, offering observability features that show how AI models are actually using enterprise data in production would help organizations optimize performance and identify new automation opportunities," he said.

Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than three decades of experience. He covers analytics and data management.