This team solves some of the most intricate and abstract technical problems that can be applied to the financial sphere. They are almost like a research lab, positioned within an extremely high pedigree and discreet trading business.
A researcher might ask whether an obscure real-world quantity can be measured at all. There are no metrics, no benchmark, no clean labels and the data arrives broken - this team has to invent the measurement. These are fascinating problems involving vast datasets and with genuine material reward and business impact.
These are deliberately not ex-finance people - physicists, astrophysicists, materials scientists and geoscientists who moved into applied research or other industries.
If you have spent your career pulling signal out of a noisy instrument, this is the same problem pointed at the economy instead.
The work is petabyte-scale real-world data: sensor readings, geolocation, imagery, unstructured event data. Iteration cycles run weeks to months, so the work is thorough rather than sprint-to-dashboard. A separate engineering team productionises the models, so you stay doing research. And you sit with the people who use your work two or three days a week, which means you find out quickly whether it mattered.
What we are looking for
- A doctorate, or a technical degree with real research behind it, in physics, geophysics, astronomy, EE, statistics, maths or similar
- Evidence you have owned a hard measurement problem end to end - the instrument, the pipeline and the model
- Strong Python ideally - other languages ok too, as long as technical capacity is there
- No finance background required, expected, or particularly relevant