We have identified common use cases, algorithms used for solving these use cases within the AWS stack and common patterns. This site is a searchable catalog of code snippets that can be used as lego blocks to build a working Machine Learning model. The goal of this project (the Pythia workflow and this website) is to provide a least resistance path for building a Machine Learning PoC.

Can there be better algorithms? better PoCs? better architectures to do the same? Yes, for sure! But we believe that this gets us to a working model faster than any of those methods since it involves generic, bite-sized code samples that you can use as lego blocks to build up your sample ML application. We would love your contribution to make this better.

Yes you can! This website is not for everyone; its not for data scientists (unless they like to search for code snippets); its for builders that can hack their way intelligently to a working, sensible PoC; it's for showing end customers the art-of-the-possible. If you like building a working PoC using the guidance we provide and code snippets, and if you have a custom dataset that is different from an example you see in [docs / blog posts / github links / medium articles] then use this. If your dataset is close to one that you see in a published example, use it and don't try to build this from scratch.

Absolutely! This is mostly open for Internal Amazon employees to contribute. We are determining ways to scale the contributions from all of you. We will post more guidance on that here. Until then, reach out to us internally, for Amazonians.

All documentation on how to contribute is available in the corresponding github repository easy-ml-pocs for this website.

Yes! This tool is free, and you only pay for what you use with services on AWS.