Microsoft Rethinks RAG for an Agentic AI World
Microsoft Ignite last week was brimming with AI news, from agents on the Windows 11 task bar to new voice commands for Copilot.
One intriguing announcement was Foundry IQ, which runs in Microsoft Azure and is currently in public preview. Microsoft says it augurs in the next generation of RAG, partly by removing the repetitive steps that begin every RAG process.
The level of abstraction in Foundry IQ is not good news for the startups that emerged in recent years to specialize in RAG or its component parts. It shows how much the AI inferencing world has changed during the past two years, and how the individual components within RAG, such as vector databases, feel less special.
A Knowledge Layer for Agents
Central to Foundry IQ is the concept of a knowledge base. It’s more than a database or data lake; a knowledge base gathers data around one topic, creating a standing library that all queries can tap—a “unified knowledge layer for agents,” as Microsoft describes it.
To access the rest of this article, you need a Futuriom CLOUD TRACKER PRO subscription — see below.
Access CLOUD TRACKER PRO
|
CLOUD TRACKER PRO Subscribers — Sign In |
