CData Software has launched Connect AI Gateway, a platform designed to manage AI agent access to enterprise data, business definitions, and model routing. The tool aims to improve agent accuracy and control costs by enforcing data permissions and filtering information before it reaches the model.
The launch of CData Connect AI Gateway addresses the critical need for governance as streaming and media enterprises deploy autonomous agents. By filtering and aggregating data before it reaches the context window, the platform can reduce token consumption and operational expenses, which CData claims varied by up to 175-fold between models in internal tests. For the streaming ecosystem, this provides a framework to integrate fragmented data silos while maintaining strict audit trails for AI-driven actions. As companies move beyond simple chatbots toward agents that execute tasks, the ability to maintain a persistent context graph outside of individual models will be essential for vendor flexibility. Watch for the upcoming automated routing feature that selects the lowest-cost model for specific prompts.
CData has launched its Connect AI Gateway in early access to provide centralized governance for enterprise AI agents. By enforcing row-level permissions and integrating business definitions from sources like Microsoft Power BI, the platform ensures data accuracy and security, while reducing operational expenses and token consumption for streaming and media enterprises.
The gateway provides a centralized platform to manage how AI agents interact with enterprise data and models, enforcing security policies and ensuring consistent business definitions.
It uses a context engine that integrates business definitions from tools like Fivetran dbt and Microsoft Power BI to prevent agents from miscalculating metrics.
Yes, the platform includes routing features that allow administrators to set token budgets and select models based on cost, potentially reducing operational expenses significantly.
Internal benchmarks showed the platform achieved 98.5% accuracy in enterprise queries, compared to 65-75% for other Model Context Protocol providers.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
