Neat, I was building something similar, but after building a few workflows in my editor, I've realized that all of them are pretty much the same: get labeled dataset, train to produce a model, use that model on some other dataset.
Seems like a suitable candidate if you ever get tired of KNIME[0].
[0]: https://www.knime.com/why-visual-workflows
alt Hacker News
Neat, I was building something similar, but after building a few workflows in my editor, I've realized that all of them are pretty much the same: get labeled dataset, train to produce a model, use that model on some other dataset.