I'm glad you can joyfully collaborate on notebooks in real-time. How is that product, or iko.ai, different from pure JupyterLab which has real-time collaboration?
In addition to real-time collaboration, iko.ai is useful for the parts described in my initial reply to this thread (experiment tracking, parametrization, model deployment, packaging, and monitoring. Using your own Kubernetes clusters on different cloud providers and using the billing account on your cloud providers, or client provisioned clusters). In other words, all the problems one faces the minute they start working on a non toy project. If someone is a teacher/student or doing work where the notebook is both the medium, the output, and the outcome itself of the work, then any real-time collaborative notebook will do as they won't need other parts. If someone does machine learning professionally (consulting, or an internal machine learning team in an organization), then they'll need the parts cited above.
Jugurtha|4 years ago
In addition to real-time collaboration, iko.ai is useful for the parts described in my initial reply to this thread (experiment tracking, parametrization, model deployment, packaging, and monitoring. Using your own Kubernetes clusters on different cloud providers and using the billing account on your cloud providers, or client provisioned clusters). In other words, all the problems one faces the minute they start working on a non toy project. If someone is a teacher/student or doing work where the notebook is both the medium, the output, and the outcome itself of the work, then any real-time collaborative notebook will do as they won't need other parts. If someone does machine learning professionally (consulting, or an internal machine learning team in an organization), then they'll need the parts cited above.