Documint is a web service for processing documents (primarily PDFs) with various actions: rendering HTML+CSS to PDF; splitting PDFs; generating thumbnails, etc.
Actions are performed within a session. A session has isolated storage for documents (both inputs and outputs) that is cleaned up once the session is terminated. This way a chain of actions may be performed without having to transfer intermediate results back and forth.
Action results are identified by a URL that may exist before the resource has any data. Actions will wait for data to exist before continuing, so it is possible to set up a processing pipeline without having to wait for each intermediate step to complete first.
See the docs for API documentation and detailed information on the HTTP API.