Open Source Software Library

Advanced Analytics Workspace (AAW) is Statistics Canada's open source platform for data science and machine learning (ML) for advanced practitioners to get their work done in an unrestricted environment made by data scientists for data scientists.

This project was designed to integrate with Statistic Canada’s The Zone platform. In The Zone, users can create individual workspaces with Kubeflow Notebooks. To persist their work, users can attach volumes to their notebooks. These volumes are not automatically deleted when a notebook is deleted. Users are free to move them or to reuse them. As a result, over time, users tend to amass several unused volumes. These volumes sit on the cloud and result in unnecessary costs. The purpose of this project is to design an automatic system that detects these unused volumes and safely removes them, reducing unnecessary expenditure. Despite being primarily designed for Statistics Canada, this project strongly values open-source practices. The codebase is available for public use and the entire development process is documented on the issues page. The volume cleaner was designed to have as little coupling as possible so it can easily integrate into projects outside Statistics Canada. Using open-source and being open-source was an important development philosophy.

Small Electron desktop application for working with Kubeflow environments. It provides a client-side wrapper that can make Kubeflow workflows easier for analysts or platform users to access.

Golang replacement for Kubeflow Jupyter web APIs used in the Statistics Canada platform environment. It provides service endpoints for notebook management in Kubernetes-based analytics workspaces.

Container images built for use with Kubeflow in Statistics Canada data-science environments. The repository supports repeatable notebook and workload images for analytics users.