Element Data

Wild Turn around

For a short period in 2019 I was brought on board to a start-up going through some transition and on shaky ground. Element Data needed to take their Proof of Concept Machine Learning technology and create an MVP design of a product suite that could sit on top of their tech stack. Additionally they needed guidance on how to shift the company from a from Technology R&D to a Product Company. After only three months working at Element, the company was dissolved because it was revealed the “special sauce” of their ML technology was not special and was in fact off the shelf technology from other companies.

  • The process to train the ML models was very manual and involved a lot of human intervention and email exchanges.

    I suggested a progress-tracking UI for ML model creation. Improving visibility compared to black box processes and email exchanges.

  • Once a customer had a model they could either connect an API or upload a data set to be scored by that ML model, this UI was built to not only show the scoring over time but also reflect the efficacy of the model itself over time and it that predicted results matched the actual scores.

  • During my brief time at Element I also redesigned their approach to data visuals and overhauled a demographics tear-away report that some customers received at the time.