You're building a new product. Before the first screen ships you already need a database, an admin to see and fix data, business logic, and eventually an API for your web and mobile clients. That's days of plumbing before you've tested a single idea.
Vantage collapses the early backend into one flow — and lets you graduate to real, served APIs when you're ready.
From idea to API🔗
The idea
A new product or feature gets the green light. You don't know the final data shape yet — and that's fine.
Describe it to Vantage
Tell your agent what you're building. It scaffolds a UI on a schemaless SurrealDB — no migrations to write before you can click around.
Shape data & logic
Tweak records, add business rules in Rhai, wire up actions, and confirm the whole thing actually works end to end.
Add a restricted role soon
Define a scoped role that exposes only what a public client should ever see.
Export the API soon
Publish that role as a standalone endpoint — your data plus your Rhai logic, served as a real backend.
Wire up your apps soon
Point your web and mobile clients at the endpoint. The admin you built and the API your users hit share one source of truth.
constructionSteps 1–3 ship today. Roles and API export are the server-side facade APIs on the roadmap — built on the Vantage Framework, which already powers facade services in Rust.
A gradual path, not a rewrite🔗
In an existing codebase, Vantage is complementary. Stand up a facade or middleware API in front of the databases you already have, move a frontend onto it one screen at a time, then delete the legacy code behind it — no risky big-bang cutover.
No hand-rolled API to keep alive — generation handles the CRUD, the logic and the serving.
Vantage is database-agnostic, so you can move to a new DB underneath without your clients noticing.
High-performance modern Rust that runs on your own Kubernetes — fully open-source and forever in your control.
Why developers pick it🔗
- Config-as-code, in your repo. No generated-and-forgotten scaffolding — the backend is YAML plus Rhai you diff in a pull request.
- Schemaless to start. SurrealDB means no migration treadmill while the data shape is still moving.
- Talks to what you already have. SQL, SurrealDB, MongoDB, GraphQL, REST — read from one backend and write through another while you migrate.
- Native and fast. GPU-accelerated, virtualised tables that stay smooth past a million rows.
See it in action🔗
The SpaceX GraphQL example points the console at a public GraphQL API — launches, rockets and capsules as browsable, drillable tables — so you can see exactly how a schema becomes a UI.