Application auto-discovery
How Kabeen AI automatically identifies the applications actually used across your organization
Overview
Kabeen AI continuously observes the web traffic collected by Kabeen agents (user agent and browser extension) and matches it against a catalog of more than 80,000 professional applications to automatically build your application inventory.
You have nothing to declare manually: the applications your team uses day to day surface on their own, qualified with their publisher, functional category and usage level.
How it works
- Kabeen agents collect the URLs visited by users (domain and path only, never the content).
- Each URL is matched against the Kabeen catalog (SaaS publishers, business software, consumer applications).
- When a URL does not match any catalog entry but shows the signals of a professional application (subdomains such as
app.,cloud.,portal.,admin., paths/app,/dashboard,/portal...), it becomes a discovery candidate. - The detected application appears in the Discovered applications panel of your inventory.
Reviewing discovered applications
From your inventory:
- Open Applications
- Click Discovered applications (top right of the list). A badge shows the number of applications awaiting review.
- For each suggested application, you can:
- Add it to your inventory (one click, with its category and publisher pre-filled)
- Reject it if it is not relevant to your organization
Rejections are remembered: a rejected application will not be suggested again.
Discovery timeline
The first applications usually appear after a few days of collection, the time agents need to gather enough usage signals to distinguish a truly used application from a one-off visit.
Best practices
- Let agents collect for at least a week before the first review: qualification stabilizes with volume.
- Revisit the panel regularly: new SaaS tools adopted by teams surface over time, with no IT intervention required.
- Combine auto-discovery with manual additions from the catalog for very recent or very specific applications.
Privacy
Auto-discovery relies only on URL metadata (domain and path). Page content, forms, credentials and keystrokes are never collected. See the Data privacy page for details.