A global shipping firm wanted to automate its process of vetting ships, to assess risk to goods in transit. We used machine learning and text data mining techniques to create an algorithm that quickly scans ship inspection reports and related shipping and customers reports and flags issues. 90% of nominated ships can now be vetted in 8 minutes instead of 2 days.

“It has been a massive help for operations to reliably plan maintenance in line with what our own data are telling us, rather than rely on manufacturer recommendations.”