Building a Machine Learning model is challenging. Tools such as the Team Data Science Process from Microsoft explains part of the problem, but they underestimate complexities outside of the model experimentation/development phase. 

A recent paper from Google titled "The Hidden Debt in Machine Learning Systems" aimed to highlight that only a very small part of developing/deploying a Machine Learning project is the creation of the model. At Advancing Analytics, we help customers build Machine Learning Models in an applied way - Hacks are great, but if a model does not offer a Return-on-Investment, then it is not a success. Over the last few years we have worked with customers capturing their requirements. Being able to capture the requirements to deliver a Machine Learning project is vital, but it is complicated. To make our lives easier, we created our Machine Learning Canvas. 

In this session we will walk through how to use the Canvas to ensure that your next project is not only successful, but that you take a holistic view and understand all the requirements to take an idea from model to production