Description: Evaluation Proxy enables local evaluation across any platform with ultra-low latency. It was in beta version for two years. Key updates: • Universal local evaluation: Supports low-latency evaluation with large cohort targeting without requiring language-specific SDKs. • Exposure tracking: Native support to auto-track exposure events.
What it is: N-Day LTV brings cumulative revenue measurement to Retention Analysis, so you can track how much value a cohort generates across fixed day windows: Day 1, Day 7, Day 14, Day 30. To use it, set a start event, pick a revenue return event, and toggle N-Day LTV under the Return Event controls.
Description: You can now connect any Slack channel as a feedback source in AI Feedback. Once connected, AI Feedback automatically aggregates and surfaces insights across all messages in that channel , no manual tagging, no copying and pasting, no context switching.
We’ve introduced a new URL Tracking plugin for the Amplitude Session Replay Browser SDK! This enhancement ensures we accurately capture client-side navigations in SPAs (Single Page Applications), a critical prerequisite for features like Jump to Key Moments.
Description: Our platform is going AI-native with the launch of 3 incredible innovations: Global Agent: a powerful AI agent that can answer data questions, create artifacts, and take action across the entire platform (also available via Slack).
Just in time for Guides and Surveys' mobile GA launch, we added the capability to copy a guide or survey to a different platform (app). Simply use the existing 'Duplicate' buttons which will now trigger a platform selection modal in case you have multiple platforms configured. Why: Customers often want to show the same message on all of their apps.
Why: Teams struggle to track and clean up old flags, which can lead to technical debt and cluttered flags. Stale status gives clear visibility into which flags need follow-up and nudges owners to take action, helping teams keep flag usage healthy and intentional.
Description: It's tough for customers to quickly customize positions, size or offsets of anchored/embedded guides and surveys from the preview on amplitude.com. They have to make changes, hit preview to open their app, close it, and then edit and make changes, repeating this loop until everything is just right. No more!
Description: Previously, you could only sort on the table columns in the syncs table, limiting your visibility of the status of your syncs. Now, with several new filters and an added column, you can get easy glances at the status of your syncs.
• If you jump into the Amplitude product and view the new Dashboard Agent run, you can see two separate deep dives: (1) summary + top overall insights like before and (2) a new section dedicated to the week-over-week differences if the agent was run the previous week.
Description: Lots of customers use AI Visibility regularly, but we've had many complaints about missing data because reports failed to generate. Report generation requires thousands of LLM calls, so there are many potential failure points. Over the past 2 months, we've made lots of improvements and reports now have close to a 100% completion rate.
Users can now set existing agents as discoverable across the org so that others can see the insights and reports being generated. All discoverable agent threads will be viewable in your agent inbox (/agents/overview page). On any Dashboard, you can see what other agents have been created and made discoverable then check out those insights.
Releasesby Glyph