CData Software introduces Connect AI Gateway, a new platform providing IT teams centralized governance over AI agents’ access to enterprise data, managing model usage, data permissions, and operational costs while improving response accuracy.
- Centralized control of AI model, tool, and agent permissions
- Context engine aligns AI outputs with corporate data definitions
- Potential for automated selection of cost-efficient AI models
What happened
CData Software has launched the Connect AI Gateway, an advanced platform designed to centralize control over artificial intelligence agents’ access to enterprise data. This new product enters early access and extends CData’s prior Connect AI platform, enabling IT teams to register AI models, tools, and agents in one place and establish detailed usage policies.
The gateway also integrates a context engine that leverages structured business data, definitions imported from analytics tools like dbt and Microsoft Power BI, and user interactions to ensure AI-generated answers remain consistent with corporate standards. Additionally, the platform records audit trails detailing which prompts, models, and policies influenced each AI response.
Why it matters
Enterprises deploying AI agents face challenges ensuring that automated responses align with internal data definitions and governance policies. Without proper controls, identical queries such as “revenue” can yield varying results depending on the AI model or data source accessed. CData’s context engine addresses this by applying standardized business logic across all AI interactions.
The Connect AI Gateway also offers IT administrators policy-based routing of requests to specific AI models, factoring in features and cost considerations. This promise of automated, intelligent model selection aims to reduce operational expenses by selecting the lowest-cost AI option without sacrificing accuracy or security.
What to watch next
CData plans further enhancements including automatic cheapest-model selection driven by intelligent analysis of prompts and system contexts, moving beyond initial policy-based controls. Expanded support for importing business definitions from broader data catalogs is also expected to ensure wider semantic alignment.
Organizations evaluating AI governance tools should monitor Connect AI Gateway’s early access outcomes, especially its effectiveness in delivering accurate, cost-efficient enterprise responses with a full audit trail. Adoption may signal a shift towards integrated, policy-driven AI governance combining data management with AI operational oversight.