CloudDM has been evolving for several years. A large part of the codebase was written by our team through traditional development practices.
Over the past six months, we increased our investment in the project and started adopting AI-assisted development more extensively. The main reason is that database systems are becoming increasingly complex — supporting many different databases requires a huge amount of domain knowledge, and AI has become a useful tool to help us explore, implement, and validate these scenarios.
Our current workflow is mainly:
AI-assisted implementation → human review → testing → iteration.
The goal is not to replace engineering judgment, but to help us move faster while keeping the quality under control.
For trying it out on a small project, we provide binaries, Docker images, and Kubernetes deployment options.
For a quick local setup, we recommend the Docker `alone` version. It includes the required metadata database, so after starting the container you only need to complete a simple initialization process.
Hi HN,
We just released CloudDM 4.0.
CloudDM started as a database management tool, but after working with database teams for years, we found that teams need more than a SQL client.
They need SQL review, collaboration, permissions, and database workflows.
CloudDM 4.0 is a complete redesign around these scenarios.
Would love to hear feedback from the HN community.
Cool How did you make it and How can I use it my simple project?
Thanks!
CloudDM has been evolving for several years. A large part of the codebase was written by our team through traditional development practices.
Over the past six months, we increased our investment in the project and started adopting AI-assisted development more extensively. The main reason is that database systems are becoming increasingly complex — supporting many different databases requires a huge amount of domain knowledge, and AI has become a useful tool to help us explore, implement, and validate these scenarios.
Our current workflow is mainly: AI-assisted implementation → human review → testing → iteration.
The goal is not to replace engineering judgment, but to help us move faster while keeping the quality under control.
For trying it out on a small project, we provide binaries, Docker images, and Kubernetes deployment options.
For a quick local setup, we recommend the Docker `alone` version. It includes the required metadata database, so after starting the container you only need to complete a simple initialization process.
The detailed installation guide is available in the project documentation: https://github.com/ClouGence/open-cdm