Level Up Your APIs: Ditch REST, Embrace GraphQL

REST has dominated web API design for a long time. However, as applications grow in complexity, REST's limitations make maintaining API and client code difficult. This talk explores why GraphQL is a better choice for building modern web APIs, and how to get up and running with a new GraphQL API service quickly. GraphQL puts more power in the hands of client developers: Improved Developer Experience: A single endpoint and a strongly typed schema means that type safety is built in during client development. Flexibility: Clients can request any combination of data in a single query Power: Libraries like Relay allow each component to declare its data needs, and the Relay compiler can automatically merge multiple query fragments into a single API request To support science data sharing and collaboration, the CZI Science engineering team has developed a pair of GraphQL-based tools that make it easy for Science teams to manage their data in the cloud, run processing workflows on their data, and produce strongly typed output: Platformics: Express your data schema as a human-readable YAML document using LinkML Expose your schema via a GraphQL API built on top of the latest-generation Async Python libraries Metadata is persisted in PostgreSQL and data is stored in S3 Generated code can be customized A recording of this session is available on YouTube: https://youtu.be/CzzYWSwb6Zw