Google built gRPC. Amazon built Smithy. Facebook built GraphQL. Palantir built Conjure.
These inventions all share a common design: schema-first automation for API development. This saves developer time and increases product velocity.
We built Fern to productize this design and make it accessible to all software companies.
Schema-first API design
Fern embraces schema-first API design: You first define your API in Fern. After that, you implement your API and generate helpful outputs, like SDKs or a Postman Collection.
With schema-first API design, you unlock a few benefits:
- Best practices built-in. Our compiler will warn you against common anti-patterns and pitfalls.
- A single source of truth. No more keeping your OpenAPI spec up-to-date with your backend. Out-of-date documentation is a problem of the past.
- A simple format encourages collaboration. An easy-to-write format means that backend and frontend devs can collaborate on API design. If you have multiple dev teams, a standardized format makes it easy for teams to learn about and consume each other’s APIs.
The Fern compiler handles the tedious and time-consuming developer work that goes into building a new API. This frees up developer time to build your product faster.
On the server side, Fern generates the types and networking logic for you. Just implement the business logic, and you’re good to go.
When consuming your new API, Fern’s auto-generated SDKs are a dream. You get client libraries to use in tests, your frontend, and other microservices. They let you build faster.
- Think in terms of operations, not HTTP. Endpoint calls are just function calls. No need to hardcode HTTP paths or append query parameters.
- Type safety to ensure you’re using the API correctly. If you’re not, you’ll get compile errors.
- Auto-completion in code editors instead of reading backend code to understand the API.
And unlike most auto-generated SDKs, ours are idiomatic and feel handwritten. You’ll be excited to show them off to your customers.