Integrate Git with Streamlit for Robust Versioning

This Cognite Hub Post outlines a workflow where development is done locally with the repository and then synchronized with CDF. Here at Celanese’s CDP (Citizen Developer Program) we have an different process/dynamic. Most of our citizens do not have access to local developer tools like VSCode, etc. Rendering the mentioned approach problematic to be applied in our use cases. A Git-based solution, similar to what Azure ML Studio: Git integration - Azure Machine Learning | Microsoft Learn offer, would ensure a more inclusive version control process. Azure ML Studio allow users to specify their Git repository details—such as the URL, branch, or even a specific commit—to connect their code base directly to the platform. Additionally, the integration could capture Git metadata (like commit hashes and messages) and associate it with each version, providing clear traceability and making it easier to track which code changes correspond to specific deployments.
The CDF would benefit from this by adopting a structured Git-based workflow. The repository would be organized with two main branches: one dedicated to versioning, where each commit represents a stable version, and another serving as a work-in-progress branch for ongoing development. Once the work-in-progress branch reaches completion, it would be merged into the main versioning branch, creating a new stable release. A built-in selector could allow users to easily switch between different versions from the main branch or continue working on the latest updates in the work-in-progress branch, ensuring flexibility and control over the development process.
With Git integration, multiple developers could collaborate efficiently, even without local development environments. The platform would maintain a consistent history of code changes, enabling better debugging and rollback capabilities while ensuring a seamless and transparent development lifecycle.