Overview of Memories in Scompler AI
Scompler AI Memory enables Scompler AI to remember information about you across chats. This way, you no longer have to repeat your role, preferred formats, or tone in every new chat.
The Memory feature is part of Scompler AI and is turned on by default for all projects and users. You can turn it off for yourself at any time, and administrators can deactivate it for the whole project in the project settings. Using the Memory feature consumes AI credits.
How Scompler AI Memory Works
A memory is a short piece of information about you, how you work, or how you want Scompler AI to behave in specific cases, for example "Writes LinkedIn posts without emojis." Each memory contains a fact and a short summary of it. The summary helps Scompler AI recall longer memories efficiently.
Scompler AI creates and maintains memories in three ways:
- On request: You ask Scompler AI to remember something.
- Automatically: After each answer, Scompler AI saves what you stated about yourself that applies beyond the current task.
- Regular clean-up: Scompler AI reviews your memories daily and merges or removes them where needed.
At the start of every chat, Scompler AI loads the memories that are relevant to your request. During the chat, it can search for further memories when needed. You can view, search, filter, and delete your memories at any time.
Your memories are personal. Your colleagues cannot see them. Sensitive information such as health data or ID and bank details is never saved.
Benefits of the Scompler AI Memory Feature
![]() |
![]() |
![]() |
||||
| Better First Drafts |
Less Repeating | Fewer Credits per Result | ||||
| Scompler AI starts every chat with your context instead of a blank sheet. Your preferences are applied from the first draft, even if your request does not mention them directly. |
You no longer have to repeat the same instructions and corrections in every new chat. Scompler AI gets to the actual work right away. |
With your context available from the start, you need fewer corrections to get the result you want. This saves time and AI credits. |
