Arvados consists of:
-* *Keep*: a petabyte-scale content-addressed distributed storage system for managing and
- storing collections of files, accessible via a variety of methods including
- Arvados APIs, WebDAV, and FUSE file system mount.
-
-* *Crunch*: a Docker-based cloud and HPC workflow engine designed providing
- strong versioning, reproducibilty, and provenance of large-scale computations.
+* *Keep*: A petabyte-scale content-addressed distributed storage
+ system for storing, managing and versioning collections of files.
+ Like git for big data. Interoperable data access by a variety of
+ methods including WebDAV, FUSE file system mount, and Arvados APIs.
+
+* *Crunch*: A container-based cloud and HPC workflow engine providing
+ strong versioning, reproducibilty, and provenance of large-scale
+ computations. Supports [Common Workflow
+ Language](https://commonwl.org) for describing workflows.
* Related services and components including a web workbench for managing files
and compute jobs, REST APIs, SDKs, and other tools.