A tour · 6 stops · 2 min
How much AI?
AI across the product range, how much space it takes, and what was still in preview.
This is the tour’s text. In 3D each stop frames the notes it talks about; every stop below links to its view.
A tour · How much AI?
How much of the release notes is AI?
Across the whole archive, Jev classifies 7.2% of notes as primarily AI. That total spans the early predictive features and Agentforce. It changes if we include AI-related work inside other capabilities, or count only what was generally available.
Jev is TypeSafe’s classification model. These counts measure documented changes, not adoption or engineering effort.
See this stop in 3DInclude changes that involve AI
AI overlaps with other work
Outside the AI category, Jev flags 648 notes as AI-related. The largest groups are analytics and service. A count based on the primary capability alone leaves them out; their colours show where that work sits in the broader AI-related band.
See this stop in 3DRead across the product sections
AI outside the AI chapter
Industries contains more AI notes than Einstein & Agentforce. Across the archive, 17% sit under the AI-branded product family. Salesforce files the rest throughout its product sections, including Service Cloud and Platform.
See this stop in 3DAvailability in the notes · Since 2024
More previews in the AI total
Jev labels 34% of AI notes in this period as planned, pilot or beta, compared with 14% of other notes. A preview announcement and a GA change each count once, despite offering different availability to a team planning an implementation.
Lifecycle describes the note’s date, not availability today. Planned changes sit at the back.
See this stop in 3DYour turn
The AI total depends on what you include
AI appears across product families, and the notes include previews alongside GA changes. The words below let you separate those views and follow a point to its source, where Salesforce gives the actual availability and licence requirements.
See this stop in 3D