Sportivision — case study

Event photographers spend hours sorting their photos, and participants scroll through hundreds of pictures to find theirs. On Sportivision you type your race bib number and an image-recognition AI finds the photos taken during your effort.

I launched it alone, from the code to the first sales: the web platform, a Python computer-vision algorithm that reads bibs and bike plates while respecting GDPR, partnerships with organisers, a press campaign and paid campaigns. Photographers upload in bulk and the AI detects the participants; organisers get better coverage and a new source of revenue.

The feedback that matters most: photographers telling me they have turned professional, because Sportivision gives them an audience and the confidence to start. On the participants' side, what touches me is seeing them post photos from the site or tell others to find theirs with us.

For the 20 km of Brussels, where a competitor was already established and Sportivision could not be present, I quickly built a tool that lets participants analyse their time and their race. Nothing to do with photos: the goal was to make the brand exist and collect email addresses. I applied my marketing class on mailing and campaign cost calculations: with €50 of ads, 1,048 people signed up, more than double my goal of 500.