Roadmap

Aspirational, subject to change based on feedback and resource availability. With that caveat, please read on ...

Q3 2015 - Initial beta

Interactive Python notebooks.
Ability to work with data to do data analysis, visualization, and transformation using Python, SQL and BigQuery
Ability to deploy DataLab as an AppEngine module in Google Cloud Platform.
Integrated with the git source repository associated with a Cloud Project for notebook management.

Move off deprecated components: git commit UI in Cloud Console and related library components to move files
Enable local run scenario: allows us to move forward without building a git client and enables user credentials for access
Move cloud deployment from App Engine Flex to GCE - halves the cost for user and improves deployment
Initial CloudML beta support

Go from three deployment scenarios to one based on feedback: maximize focus, address pain-point of requiring local Docker installation. Focus will be on running notebook server and kernels in GCE
Streamlined cloud deployment: command line tool bundled with Cloud SDK
Bundled git client experience - likely using ungit
Automatic backups of notebooks from the persistent disk
Update libraries to match recent developments: Google SQL support in BigQuery
Dataproc-based deployment - through initialization actions
Significantly up-leveled CloudML experience to support code-free and code-full usage. Starting with images and structured data toolbox support. This is one of the largest areas of current and future investment

(More broad brush strokes, timeline TBD)

Datalab as a service: walk up and use experience for getting started
Enable Drive-based collaboration experience: commenting and easier sharing. This will require Drive to be enabled with the appropriate account. No backend needed for browsing/editing. Code execution will require a VM
Simpler backend acquisition from Datalab UI
Further up-leveling of CloudML experience with significant toolbox additions for more scenarios
A data-first experience