Go to main navigation Navigation menu Skip navigation Home page Search

Kristoffer Milonas

Meet Kristoffer – Staff Research Scientist at Upstart and PhD in Finance alum.

1. Describe your role and what it is that you do overall and on a day-to-day basis. 

The company develops models based on AI and machine learning that decide whether individuals get loans and at what rate. My main responsibility is to prevent these models from having unintended consequences, both in terms of potential risks that are not well covered by the models, and in terms of fairness between demographic groups. In addition, I answer questions about our models from our bank partners (who are the ones ultimately making the loans based on our models) and support external model validations. The "research" part of the job title is there largely because model risk inherently involves articulating and answering new questions, and because traditional fairness and model risk metrics need to be adapted to our context. On a day to day level, that translates into designing economic and statistical analyses, collecting data, running regressions and simulations, and running meetings on how to make these analyses as impactful as possible.


2. What interested you about the field you are currently in? 

I enjoy applying the quantitative and business skills I gained from my education to new questions where there's often not a clear roadmap for how to do things. The mission of the company also resonates with me - far too many people can't get credit at all or pay way more than they should, partly due to lenders using outdated models. 


3. Why did you choose to study your subject area at SSE? 

I enjoyed the undergrad finance courses, it seemed natural to continue!


4. How did your time/education at SSE help guide you to the career journey you have embarked on? 

In many ways! Our work sometimes raises pretty fundamental economic questions, and the grounding in economic theory has been useful for helping me structure a framework for addressing them. Aside from the obvious benefit of the quantitative courses, I also found that some of the courses taught you how to integrate ambiguous information and situations where it's not clear from the beginning what the most material questions are. I have found this very useful for real world settings where it's much less clear than in the textbook what the right questions are, let alone the right answers.

 

5. What path did you take from graduation to where you are now? 

  • PhD at SSE, during which I also was a visiting researcher at LSE.
  • Research Economist at the Bank of England, doing a mix of research and policy work.
  • Associate Director at Moody's Analytics, developing predictive models for credit risk applications.