Repo-first platforms outpace Supabase
CTO at insurtech startup on how AI code generation undermined Supabase's core value proposition
The real advantage in Firebase and AppSync is that the app can be treated like one auditable software project instead of a mix of dashboard state and scattered backend logic. In practice that means security rules, schema definitions, deployment settings, tests, and application code can all live as files in the repo, so a CTO can inspect a pull request and see what changed. With Supabase, some of that can live in git through the CLI, but the working default is looser and easier for AI generated projects to fragment.
-
Firebase makes this concrete by letting teams define Firestore, Realtime Database, and Storage rules in local rules files referenced from firebase.json, then test them in the Local Emulator Suite. That creates a normal code review loop where auth and data access changes are visible next to product code.
-
Amplify and AppSync push even further toward repo driven backends. Amplify Gen 2 provisions backend infrastructure in TypeScript code, and Amplify Hosting is designed to connect builds to a Git branch, so infra and app changes move through the same commit history.
-
Supabase does support git based workflows through supabase init, config.toml, and migration files under supabase/migrations. The issue is not that git is impossible, it is that teams often still end up with database structure, policies, and platform settings split across SQL, config, and console workflows, which makes AI built apps harder to read end to end.
This points toward a split market. Professional teams will keep favoring stacks where repo review is the default unit of control, while Supabase remains strongest where speed matters more than auditability. As AI writes more of the software, the platforms that win will be the ones that make every important change show up clearly in git.