I have spent the last 12 years working on credit risk and machine learning models inside major European banks. The work has two sides: building the models (IRB PD/LGD/EAD, IFRS 9 expected credit loss, stress testing and scenario analysis), and auditing how they are used and whether they hold up to scrutiny from developers, model validators and senior risk managers. A recent piece I am proud of is a climate risk credit model for mortgage portfolios, in a domain where the methodology was still being figured out.

Alongside the day job, I like guiding people and sharing my experience. Banking regulation, machine learning techniques and their combinations are full of useful ideas trapped behind dense documents and unclear notation for readers who are not pure quants, and I enjoy the work of pulling them out and explaining them in plain language. That is what my Notes try to do, one concept at a time. I am still learning how to do this well, and I am grateful to the readers who push back when something is not clear.

Away from the desk, I cycle. I raced at a competitive level for several years, and that period taught me more about discipline than any office has: the daily training on days you do not want to get dressed and go out, and the patience of watching small improvements add up over months.

This is what cycling has taught me: a clear goal, pursued patiently day after day, almost always pays back.