CData Software Inc. today announced three products aimed at simplifying access to enterprise data for developers building artificial intelligence applications.

They include a free Connect AI Developer Edition, an open-source Python software development kit and a new command-line interface tool. The launch addresses what the company said is the growing challenge developers face to easily access governed enterprise data for use with large language models without extensive involvement from information technology teams.

ConnectAI provides real-time access to more than 350 enterprise data sources using the Model Context Protocol while handling authentication, API management and data governance behind the scenes.

The new offerings extend the platform to make enterprise data accessible through the tools developers already use, according to CData Chief Product and Technology Officer Raviv Levi.

“Developers need the same data infrastructure that IT governs and business teams depend on,” he said. “These releases extend that infrastructure to wherever developers work: terminal, Python environment or [integrated development environment].”

The centerpiece of the announcement is Connect AI Developer Edition, a free version of the company’s platform that includes MCP server support, user authentication passthrough, query logging and management capabilities. The product is designed to work with AI development environments such as Anthropic PBC’s Claude Code, OpenAI LLC’s Codex, Anysphere Inc.’s Cursor and the open-source LangChain.

Jerod Johnson, CData’s director of technology evangelism, said AI-assisted development is amplifying longstanding data integration challenges.

“All of the problems developers have had dealing with data still exist, but now they’re amplified because they’re using AI coding,” Johnson said. “They can get things out faster, but those old problems still stick around.”

Authentication management, application programming interface drift, schema changes, rate limits and pagination become more difficult when large numbers of AI agents and coding assistants are interacting with enterprise systems simultaneously, he said.

‘Confident but corrupt’

CData gave the example of schema changes inside business applications. AI systems may produce inaccurate answers if they don’t recognize when new fields or objects are added.

“You get confident but corrupt output,” Johnson said. “The AI agent doesn’t know that that custom object or custom field exists.”

CData said its platform addresses this issue through dynamic schema discovery, automatically detecting changes in connected systems and exposing updated metadata to AI tools.

The open-source Python software development kit is built on DB-API 2.0, standard specification that ensures consistency across different Python database drivers. It allows developers to access Connect AI through familiar Python workflows and tools such as pandas and SQLAlchemy without learning new interfaces.

CData CLI, is a command-line tool that enables developers to build and test integrations using CData’s connectors. The company says the tool is optimized for AI-assisted development environments while allowing applications to be deployed without requiring MCP infrastructure or AI dependencies.

The move is a shift in CData’s strategy. The company has historically sold connectivity technologies to developers and enterprise IT teams, but the new products are intended to create a developer-focused entry point into the same platform.

“We’re releasing products that meet modern developers, where they are: in the terminal, the IDE and open-source Python environments that are powered or backed by LLMs in some capacity,” Johnson said.

CData said data access, governance and integration are becoming as important as underlying AI models as organizations move from isolated experiments to agentic systems operating across multiple business applications. It’s betting that simplifying those layers will help enterprises scale AI development without sacrificing control.

Photo: Unsplash