Remember the user, not the session
Keep preferences, projects, writing style, recurring decisions, and trusted sources available after a chat window closes.
AI memory
Personal and team memory that follows the user across tools.
This page is about Recall Core as the user's durable memory layer: preferences, decisions, projects, notes, and long-running context that should not disappear when an AI surface changes.
Re:Yi official sitePublic content, local custody.Recall Core data stays outside the online service by default.
Work across personal and team tools
Recall Core should support one person's private memory as well as shared team context without turning every app into a separate silo.
Let people correct the record
Useful memory needs visible sources, recoverable mistakes, and a clear path when the system remembers something outdated or wrong.
Keep custody understandable
Local-first storage, explicit connectors, and Control Center ownership make it clear where memory lives and how to stop it cleanly.
Keep unfinished work visible
Recall Core can track unresolved commitments, dependencies, questions, and story threads, then surface a reasoned reminder without pretending an uncertain item is complete.
Unfinished work and foreshadowing
Open loops remain inspectable, scoped, and reversible.
Recall Core owns detection, evidence, reminders, resolution, and history. Re:Yi Control Center is where a person reviews candidates, changes status, repairs storage, and backs up or restores this local data.
Privacy boundary
The Re:Yi online service never receives open-loop content.
The official site explains the capability and may publish account or release metadata. It does not read, store, log, back up, export, resolve, or delete a user's open-loop records, evidence, reminders, or local memory content.
User path
Memory should feel editable, recoverable, and owned.
A person should be able to start small, connect one trusted source, inspect what Recall Core can access, and correct the system without needing to understand backend modules or Admin controls.
- 01Install Control Center
- 02Choose trusted sources
- 03Connect AI tools deliberately
- 04Review, correct, or remove memory
Different from agent workspaces
This is the human-centered memory path.
Agent workspaces focus on durable task context for autonomous or semi-autonomous workers. AI memory starts from the person or team who owns the long-term context.