# The AI now remembers context about your users

**Summary:** The AI remembers context about end-users across sessions, a new agentic content library is in beta, and you can configure in-app notifications in lifecycle stages.

**In short:** A more personal AI that remembers what matters.

**Published:** 2026-05-18

- You can now __configure in-app notifications__ directly within lifecycle stages, allowing the AI to proactively surface relevant messages at the right moment.
- A new __agentic content library__ is now available in beta. It can automatically build a structured, wiki-style knowledge base by processing your documents through AI, making content management smarter and more automated. (Let us know if you're interested in trying it out!)
- The __AI now remembers context about end users__ across sessions and environments. You can see this on the user details page (reached via conversation details) in the new Memory block. Additionally, if a user escalates a ticket, this information will also be included in the ticket.

__Bug Fixes__
- The AI component no longer incorrectly lowercases capitalized words (such as proper nouns) when displaying questions.
- Resolved errors that were affecting the insights aggregation feature.
- The settings icon in the sidebar now shows a clearly visible highlight when you hover over it.
- Analyzing flagged responses from the quality dashboard is working correctly again.
- The Quality Control Report now accurately evaluates questions that the AI is able to answer in regular conversations.
