In 2012 I departed to Vancouver Canada. In my mind that was the last chance to do good international experience before settling into a career. So I used a Working Holiday Visa to explore British Columbia. Some might call that a sabbatical. At that time I was very unaware of fashion trends and quite uninterested to impress with my clothing. It came to a surprise to myself when I stopped and scrutinized the Fluevog store in Vancouver downtown. Even more distressing was my interest for a pair of classic and colourful shoes. That was an unusual interest on my side. The day moved on, and so did the months. When I got back in my home country I took lessons form a local cobbler and made my first replica of these unforgotten shoes. It was an awful replica, nothing like the original, however it opened up my interest for crafting shoes. The first picture from the left is my first shoe, the second and third picture are my second attempt.
Not long after that first attempt the opportunity to build a company in Vancouver presented itself. FTSY (the name of the company) had the mission to solve the issue of returns over online shoe sale. At the time, 40% of shoes sold over Amazon.com where returned for wrong size. It was and still is, a huge financial and environmental cost. FTSY developed a software allowing the user to take pictures of their feet (using their smartphones) and a 3D model was generated and compared to a pool of shoes to guide the user towards the right model and size.
In this venture my skillset came to use under research and development. I knew how shoes where made and understood the limitations of such traditional craft. Craft that did not evolve significantly even in a mass production context. I understood the nature of fit in the shoe space (the same nature belong to the apparel space and related products). In this favourable position, I led the research into machine learning algorithms of 3D meshes and developed categorizers for shoe fit. This process, initially took the shape of a massive data collection. I interviewed hundreds of people and tested their fit preference and collected a 3D model of their foot anatomy. Than the research shifted towards testing various model to make accurate predictions.
Thanks to this venture, I lost all my reverent attitude toward AI and machine learning. Once again, knowledge is power. I now understand the limitations of this tool and I quickly understand when it is misused or and when costs connected to its implementation are under/over estimated. This piece of knowledge has been also useful as a mentor while helping start-ups into achieve their goals.
FTSY has been recently sold.
Comments