Background Tasks and Notifications
Some of Scholarly's AI creations take a few minutes — video lectures, podcasts, slide decks, deep research reports, and large flashcard decks all involve real work behind the scenes. You don't have to watch the spinner. Close the window, switch to another task, even leave the site — Scholarly keeps generating in the background and lets you know when each item is ready.
What Runs in the Background
These creations keep running after you close the create window:
- AI Flashcards — large decks generated from PDFs, audio, or research.
- AI Podcasts — script writing + multi-voice audio.
- AI Video Lectures — script, animation, narration, captions.
- AI Slides — full PDF slide deck with layout checks.
- Deep Research — multi-step web research with citations.
- Recordings — transcription after upload or live capture.
A small background tasks strip appears on your home page whenever something is generating. It shows what's running, how far along it is, and lets you click straight into a task — even one that's still processing.
Starting a Task and Walking Away
The pattern is the same across every creation flow:
- Open the create window for the feature you want.
- Pick your source material and customize the settings.
- Click Create (or Generate) — the task starts.
- Close the window (the X in the corner) or navigate anywhere else in the app.
- Carry on with whatever else you wanted to do.
The task runs to completion on its own. Closing the window doesn't cancel it — you have to explicitly cancel from the task row or the content page itself if you really want to stop one.
How You Know When It's Done
You'll see the completion two ways:
- In-app — the task row on the home page flips from "Processing" to "Ready," and the new item appears at the top of your Recents. The sidebar shows recently completed items too.
- By email — for podcasts, video lectures, and AI Slides, Scholarly sends a completion email so you don't have to keep checking. Toggle these notifications on or off in Settings → Notifications.
Email notifications are most useful for video lectures and long podcasts, which can take 10–25 minutes. Smaller tasks like flashcard generation usually finish before you'd think to check email.
If a Task Fails
A task can fail for a handful of reasons — a model being temporarily unavailable, a source file that turns out to be unreadable, an unusually long generation that hits a timeout. When that happens:
- The home page row turns into a clear error with a short explanation of what went wrong.
- A Retry button is right there — click it to re-run with the same settings.
- Failed creations do not count against your daily AI creation limit — your daily budget is reserved only for successful items.
If a retry keeps failing, swap to a different AI model (in the create window) or try a smaller / cleaner source file. If you're stuck, see Feedback and Bug Reports.
Multiple Tasks at Once
You can run several background tasks side by side. Start a video lecture, kick off a podcast, queue a research report — all three generate at the same time and appear separately on the home page. The only hard cap is your daily AI creation limit (see Plans and Limits).
Tips
- Treat the home page as a control tower. Glance at the background tasks strip when you come back — finished items wait at the top, in-progress ones show their stage.
- Pin a task as soon as it's done. That keeps it from getting buried as you create more.
- Don't worry about your battery or browser. Generation happens on Scholarly's servers, not your device. Close the laptop and the task still finishes.
Related
- Home and Library — where background tasks appear on home.
- Plans and Limits — your shared daily AI creation budget.
- Feedback and Bug Reports — what to do when a task keeps failing.