Kit Burgess

Despite their immense size, the multi-trillion-dollar energy commodity markets lag behind others in the use of big data to drive efficiency.

Harnessing live vessel tracking data enables modelling of oil flows & local fluctuations of regional tanker availability. In turn, this can be used to predict the impact on the commodity and freight markets.

Vortexa uses billions of satellite & commercial datapoints to track 99% of all seaborne oil & refined products movements, globally. Generating complex loading/discharge events and ship-to-ship oil transfers from multiple disparate data feeds in real-time is a significant Data Science and Engineering challenge. This talk discusses the ML methods used to process live AIS (automatic identification system) feeds, model vessel events, and generate real-time destination predictions & diversions. The talk also covers predicting onboard products & grades, handling contextual anomalies across multiple shipbroker/port-agent data sources, modelling freight availability & tanker supply/demand, and future cargo loading predictions.

This industry application of AI provides transparency into current tanker movements, and reduces the opacity found in the commodity & freight markets.

Biography:

Kit Burgess builds tools using probability theory. A Cambridge mathematician, he designs production Machine Learning models, relying on a rigorous statistical background to model real world systems. As one of Vortexa’s first Data Scientists, he helped the AI start-up grow to achieve an unparalleled level of transparency into oil market flows.

Before Vortexa, Kit worked in The Hut Group’s AI Incubator, developing e-commerce site personalisation algorithms.