Searching for Fashion – a Machine Learning Saga!

Product development in an ML world presents its own set of challenges and opportunities. The presentation aims to provide a walkthrough of the journey my team went through during the development of Search – a ML driven product. While doing so, I aim to highlight the shortcomings and challenges that we faced and the learnings we generated.
The product was finally rolled out across 16 different countries in Europe with each offering their unique set of languages, contexts and cultures.
The presentation also aims to expose an approach to develop a ML product in an iterative fashion which can be challenging given ML’s origin in research where time is not always the key consideration.
Lastly, the idea is not to preach or teach but rather share knowledge and let the audience decide the best way forward.