On a day-to-day basis, I use Code Ocean to run, validate, and share code in a fully reproducible environment, mostly when working with data analysis and research-oriented code, where results need to be reproducible.

One of the main use cases for Code Ocean was for reproducible research and secure code execution, along with collaboration. I used it for packaging code, datasets, and environment configurations so results can be reproduced exactly by others, making it very useful for peer reviews and validation.

Code Ocean has resulted in an overall improvement in collaboration reliability in my organization, helping us improve reproducibility and audit requirements, which are essential for some of our research-heavy or regulated workflows or tasks, and has also shortened review cycles and increased confidence in the shared results.

Our review cycles have been reduced by up to 20%, and while some improvements cannot be measured in metrics, the overall reproducibility and audit requirements have also been improved, allowing us to spend less time debugging environment issues and more time focusing on analysis and results.

Overall, there is improvement in our return on investment. We don't have to go through all the long review cycles, and most of our extra efforts involved in managing access and improving environment consistency have been reduced, which has removed excess efforts that we needed to put in and allowed us to spend less time debugging environment issues.

The best feature of Code Ocean is the compute capsule concept, which bundles code, data, dependencies, and instructions into a single reproducible unit, allowing for one-click execution. This enables anyone with access to rerun experiments, code runs, and pipelines without worrying about setup, versioning, and tracking, which are also very valuable.

The capsule concept and versioning in Code Ocean improve collaboration reliability significantly, allowing teams to spend less time debugging environment issues and more time focusing on analysis and results.

There is not much to dislike about Code Ocean, but I think the compute resources can sometimes be limited for very large or long-running workloads, and more flexible options or scaling compute may be beneficial.

I chose eight because Code Ocean can still be made a bit better. There is not much to dislike, but it lacks some flexibility for heavier jobs, and the compute resources can be difficult to manage for larger workloads.

I have been using Code Ocean for almost two years.

Code Ocean is quite stable.

Code Ocean's scalability meets all our requirements.

The customer support for Code Ocean is good. I have not needed to reach out much, but it is good.

Positive

We did not use a different solution before.

Overall, there is improvement in our return on investment. We don't have to go through all the long review cycles, and most of our extra efforts involved in managing access and improving environment consistency have been reduced, which has removed excess efforts that we needed to put in and allowed us to spend less time debugging environment issues.

I am not entirely sure if we evaluated other options before choosing Code Ocean.

I would highly recommend Code Ocean, as it is not something optional. Once you start using it, you will love it. Code Ocean is an excellent platform for collaborative and reproducible computing, particularly if your work involves sharing code, data, and results in a reliable and auditable way. I gave this product a rating of 8.