Sign-ups are failing and auth looks fine. Polylane finds the layer that actually broke.
Polylane puts your database, auth, storage, and edge functions in the same graph as the apps that depend on them, and reads the logs Supabase already keeps.
Join the waitlistEvery supported resource.
The Supabase resource types that show up in your graph.
Project Postgres Database Connection Pooler Edge Function Database Branch Storage Bucket Auth Service Storage Service Realtime Service REST API Custom Domain Secret Network Restrictions Every service in the project, watched.
Agents map the project, read its logs, and investigate what breaks.
The whole project in one graph
Every project with its databases, branches, edge functions, and buckets, plus the auth, storage, and realtime services behind them. Each resource carries its edges and its history, next to everything else you run.
Linked to the apps that use them
Polylane matches Supabase hosts in env vars across your other clouds, so the graph already knows which service talks to which project. When checkout breaks, the project is in the blast radius.
Detection tuned to Supabase's signals
Agents watch edge 5xx rates, Postgres error logs, auth failures, function errors, and service health against how each project normally behaves. No thresholds to tune.
Six log sources, actually read
Edge, function, Postgres, auth, storage, and realtime logs join investigations as evidence, alongside the security and performance advisors Supabase already runs on your projects.
Your rows stay yours
Polylane reads the Supabase Management API and nothing else: it never connects to your databases, never sees your rows, and never stores project API keys. Credentials are encrypted before they're stored and the agent never sees them.
Investigations that end in a PR
When the root cause is a bad migration, a misconfigured function, or a line of code, the fix lands on the repo behind the app, evidence trail attached.
Questions.
Is Polylane an AI SRE for Supabase?
It's what most people mean by the term: agents that keep your projects, edge functions, branches and storage in one context graph, watch them, investigate issues and fix what they can, with a human approving anything that changes production. We call it self-operating software rather than an AI SRE: the point is that the project looks after itself, not that a bot joins the on-call rotation.
How does Polylane connect to my Supabase organization?
With OAuth or a personal access token. Click connect and approve in Supabase, or paste a token. Either way the credential is encrypted before it's stored, and the agent never sees it.
What Supabase resources does Polylane sync?
Projects, Postgres databases, connection poolers, database branches, edge functions, storage buckets, the auth, storage, and realtime services, REST APIs, custom domains, secrets, and network restrictions. The full list is on this page.
Can Polylane change my Supabase projects?
Agents investigate and explain. Code changes only ever land as pull requests on the repo behind the app, and any write action pauses for your approval first, with the exact request and reason on screen.
How does detection work without alert rules?
Agents read each project's logs and health signals on a cadence and judge them against how that project normally behaves: edge 5xx rates, Postgres errors, auth failures, function errors, request volume, service health, and security advisor findings. Repeat detections fold into one issue instead of re-paging you.
Does it work with my observability provider?
Yes. Datadog, Sentry, Honeycomb, Axiom, and Grafana Cloud join the graph as first-class sources, and their alerts get triaged by an agent the moment they fire.
What does it cost?
Pricing is public and self-serve: a free plan, paid plans from $80 a month, and pay-as-you-go overage past the included credits. Every plan and limit is published at polylane.com/pricing.