Doug Hudgeon

Hi Gibran, I've never used Applitools. I hope it meets your requirements.

Hi Michael, I've used Dataiku and really like it. Congrats on a great product. I don't see Dataiku losing a ton of business to Sagemaker. AWS is great at making foundational tools and not so good at making the next level up in usability. The target market for Sagemaker is the do-it-yourself company that wants to build their own tools without having to manage infrastructure.

Hi Gibran, I've never worked with WEKA but I've heard good things about it. For automation testing, what types of problems are you looking to solve with ML? And regarding using a cloud service (such as Sagemaker) vs self-managed environment, the deciding factor for me is whether I want to deploy the model in a production environment or not. If I do, then I'll reach for a cloud service because I find it an inconvenience to manage my own infrastructure in a production environment (given that all of the other infrastructure I manage is serverless).

Doug

Hi Sundar,

I really like the tools coming out of Google and Azure, and there are a number of new competitors like Deepnote on the horizon: https://www.deepnote.com/

One of the main reasons Sagemaker is the tool of choice for our book is that has the right mix of hiding the complexity of setting up servers whilst still exposing the code required to create and train models. The other reason is that Sagemaker makes it easy to deploy models to production so you can use the models you build.

The good thing about these serverless notebook-based tools is that once you know how to use one, you can easily pick up any of the others.

Doug