I'm a third-year PhD student in Econometrics and Quantitative Economics, working in Finance, Public Finance, and IO. I applied to Amazon pretty late this year, and since I was extremely busy this semester, I took the last possible deadline at every stage. I finally got the email confirming I'd passed the final technical interview on April 10, and I'm now officially in the team-matching phase.
I went with the reduced-form causal inference track, though I'm actually just as comfortable with structural IO and macro finance forecasting. Since I've mostly been working with micro data recently, I picked the causal inference track this year, and that's also mainly what the internship work would involve.
Right now I'm a short-term consultant at the World Bank through early May, on part-time CPT. I can start mid-May and work through August 30. Within the causal inference track, I'm especially interested in double machine learning, and in using text data — like user addresses — for policy analysis. I'm open to device, health, video, or basically any team, and I'm especially interested in projects that cross over into macro finance or IO. I'm very comfortable with the standard causal inference toolkit — DID, PSM, synthetic control, IV-GMM — I've studied and used all of them extensively. I'm proficient in Stata, R, and Python, and I've also picked up some MySQL.
As for the interview itself: in the first round there was a behavioral question about something I'd done outside my normal job responsibilities, and whether I'd ever missed a deadline. Then there was a domain-knowledge question where they gave me a work scenario — mine was about the effect of a prescription-refill subscription reminder on user experience, and I had to walk through the approach from DID to PSM. You really need to be clear on the specific technical details for that one. I'll write up the second round separately later.
Two books I'd recommend: Mostly Harmless Econometrics, and Causal Inference: The Mixtape by Scott Cunningham.
Discussion
Loading comments…