Ask HN: Which AI/ML software stack is quickest way from idea to ML provision?

Enterprise use a lot of tooling such as DataRobot, H2O Driveless AI, Alteryx, KNIME, Datameer, Ayasdi, Quantexa etc.<p>There is even a large list (for Big Data Ecosystem) maintained by Matt Turck which has a section for AI&#x2F;ML<p>http:&#x2F;&#x2F;mattturck.com&#x2F;wp-content&#x2F;uploads&#x2F;2019&#x2F;07&#x2F;2019_Matt_Turck_Big_Data_Landscape_Final_Fullsize.png<p>https:&#x2F;&#x2F;mattturck.com&#x2F;data2019&#x2F;<p>Cloud providers are also making significant attempts on capturing this space. - https:&#x2F;&#x2F;aws.amazon.com&#x2F;machine-learning&#x2F; - https:&#x2F;&#x2F;cloud.google.com&#x2F;products&#x2F;ai&#x2F; - https:&#x2F;&#x2F;azure.microsoft.com&#x2F;en-gb&#x2F;overview&#x2F;ai-platform&#x2F;<p>There are tons of AI&#x2F;ML open source solutions as well. Tensorflow, Keras, PyTorch, MXNet, Kubeflow and the list goes on... https:&#x2F;&#x2F;github.com&#x2F;topics&#x2F;machine-learning?o=desc&amp;s=stars<p>There is also a huge list maintained by kdnuggets. https:&#x2F;&#x2F;www.kdnuggets.com&#x2F;companies&#x2F;products.html<p>The use cases for ML are vast and every solution in the market is trying to find a unique corner for itself.<p>What do you think is the best way &amp; tech stack to go from Idea to Analytics using AI&#x2F;ML?<p>Do you have any recommendations for a decent commercial solution even if it is closed source.