File Format

A file format is a convention for encoding and structuring information so that it can be written to disk, moved between systems, and read back the same way by whoever opens it next. Formats range from human-readable text like JSON, YAML, CSV, and XML to compact binary encodings, and each one carries an implicit contract about how the bytes map to meaning. Almost everything I work with in the API space — specifications, payloads, and data exports — is a file format underneath.

• A serialization contract - Rules for how structured data is turned into a byte stream and parsed back, so both ends agree on what the data means.
• Human-readable vs. binary - Text formats like JSON and YAML optimize for legibility and diffability, while binary formats optimize for size and speed.
• Extensions and media types - A .json suffix or an application/json content type tells tools which parser to reach for.
• Schema and validation - Many formats pair with a schema so the shape of a file can be checked before it is trusted.

In real API operations, file formats are the substrate for everything from the OpenAPI and APIs.json documents that describe an API to the fixed-format records still flowing out of legacy financial systems. Getting the format right — and declaring it honestly with a media type — is what lets pipelines, gateways, and now AI agents ingest data without guessing. As the agentic turn accelerates, machine-readable, well-specified file formats are what make an API legible to a model rather than an opaque blob.