Understand Proactive
What Is Proactive AI?
Most AI models and agents still depend on an input box. You provide the context first, such as the background of a recent project or the topic you want to write or research, and only then does the AI produce a result. We call this passive AI. Proactive AI means the system can offer useful suggestions before you explicitly provide context or assign a task. AirJelly can do this after it has accumulated enough context: it can notice recent work, interrupted tasks, or follow-ups that may need attention before you ask for them.
When Does AirJelly Suggest Something?
Typical scenarios include: a follow-up appears after a meeting, someone asks you to handle something later in chat, a cross-day project needs to keep moving, repeated work should become a reusable workflow, or you are repeatedly turning similar information into similar outputs.
Where Are Ignored, Dismissed, Or Missed Proactive Suggestions Saved?
- Open Proactive Assist in Chat to collect and revisit suggestions AirJelly made earlier but you did not run at the moment.
- Open Proactive Assist on Home to review suggestions connected to your current day and recent context.
How To Reduce Interruptions
If suggestions feel too frequent, lower notification intensity, ignore a suggestion type, or turn off Proactive Assist in Settings. You can also keep only certain types of proactive suggestions enabled. AirJelly should work with your rhythm, not interrupt it.
What If A Suggestion Is Wrong Or Too Frequent?
- Dismiss suggestions that do not match your real intent so AirJelly has a clearer signal about what is useful.
- If a suggestion appears after the task is already done, check whether the related task, schedule, or reminder should be marked complete or adjusted.
- If suggestions are often noisy, reduce proactive modes in Settings and keep only the categories that match your current workflow.
- For sensitive work, pause Smart Capture or exclude the app before AirJelly turns that moment into reusable context.