It has a ton of integrations to other cloud platforms, including AWS and GCP. There is an awesome documentation portal and tutorial library for beginners. However, the biggest drawback of the platform is the pricing. You have to pay for every single server and computer that you use, which can be costly if you're scaling up quickly. It has all the basic functions that you would expect from a machine learning toolkit. Data exploration, feature engineering, modeling, etc. We use Spell extensively and it’s a good option if you want to try machine learning without the hassle of building your own machine or paying too much money for Amazon instances. It’s pretty powerful and offers excellent performance when it comes to running tasks. If you want something more enterprise ready though, you may need to look elsewhere.