What your AI no longer has to write
Your AI asks for these in a few lines ofgliss.yaml, and Gliss provides them:
Sign-in
Your company accounts, enforced in front of the app. The app writes no login code.
A database
Postgres with backups, plus a cache and object storage when the app needs them.
Secrets
Set by the person who holds them, through a link. Your AI never sees a value.
A web address
A clean address for the live app, and private addresses while you build.
Previews
A private Preview while you build, a Beta to share, then Publish to go Live.
Monitoring and upgrades
Logs, metrics, diagnoses your AI can act on, and a platform that keeps itself current.
Who these docs are for
Builders
You have an idea for an app and an AI tool. Start with the quickstart.
IT admins
You decide what apps may use, and approve what goes live. Start with the security model.
Engineers
You want more control: several services, your own tests, your own charts.
How Gliss is different
- No tickets, no cloud console, no keys pasted into chat. Build in a private Preview, share a Beta link with colleagues, then press Publish.
- IT approves it, because it was built for them. Apps run in your company’s AWS account, sign in through your company’s accounts, and stay inside the menu IT publishes. Every app has an owner, an audit trail and a cost.
- Apps stay alive. Bots watch every app and the platform, propose fixes and keep everything upgraded.