Database, Analytics, and AI/ML Roadmap – Overall Trends

by
Andrew Snodgrass

by
Andrew Snodgrass

Andrew analyzes and writes about Microsoft's data management, business intelligence, and machine learning solutions, as well as aspects of licensing... more

This section discusses high-level trends that Directions sees in the areas of Database, Analytics, AI, and Machine Learning offerings.

Near Term (2026–2027)

In the short term, Directions expects the following:

AI agents as the focus across all products. Microsoft is positioning its data products as the foundation for its Copilot and agent offerings. Fabric data agents, Fabric IQ, and operations agents reached general availability in 2026, Power BI semantic models now ground Microsoft 365 Copilot answers, and SQL Server 2025 added vector search and AI functions inside the database. However, much of this work remains in preview, including the Fabric IQ ontology and approximate vector indexing in SQL Server. Customers should expect Microsoft to keep promoting agents and should review defaults and test answers before relying on AI.

Strong push for Fabric adoption. Microsoft continues to retire features in older products and offer replacements only in Fabric. Power BI Premium P capacities must move to Fabric at each customer’s Enterprise Agreement renewal, Power BI Q&A stops working in Feb. 2027, Dataflow Gen1 is deprecated, and Power BI real-time streaming and the Stream Analytics Power BI output both end Oct. 31, 2027. The Microsoft replacements are Fabric-based and consume Fabric capacity, turning what were included Power BI features into new costs. Synapse Analytics and Azure Data Factory remain supported, but new features and migration tooling go only to Fabric.