Why use Deep Recall?
Keep useful context available as your conversations and tools change.
Spend less time repeating yourself
Section titled “Spend less time repeating yourself”Preferences, work context, and recurring constraints can matter in many conversations. Saving the useful parts gives a connected assistant somewhere to look before asking you to explain them again.
For example, a travel assistant can recall your preference for direct flights, while a planning assistant can find the dates you have already committed to. Temporary context can be useful too; include the dates that make it meaningful.
Switch assistants with your context
Section titled “Switch assistants with your context”A memory saved through one assistant can be found by another connected to the same brain. Choose a tool for the task without rebuilding all of your personal context there.
Provider support and account eligibility vary. The connection guides explain how to check what your assistant supports.
Pick up a project where you left it
Section titled “Pick up a project where you left it”Save decisions with their reasons and dates. A later conversation can recall the relevant constraints before proposing another approach.
Useful: “For the November launch, we chose a simple email invitation because people already know that flow.” Less useful: a transcript full of discarded ideas without saying which decision was adopted.
Find details in your sources
Section titled “Find details in your sources”Supported uploads and forwarded emails can create source-linked memories. Ask about a detail, then return to the document or message that supports it. Check processing and extraction results instead of assuming every file was fully understood.
Keep human control
Section titled “Keep human control”The app gives you ways to inspect, pin, archive, restore, delete, and export memories. When a plan changes, you can correct the context your assistants use.
Deep Recall helps with access to saved context. It does not guarantee automatic saving, complete recall, or correct answers. Start with a small real workflow and check its sources.