A recruiter reached out to me to schedule a brief screening call. This was followed by: 1) a 45 minute first interview with a current employee on product sense and SQL; 2) a day onsite with 4 components: 1) product interview (can you reason about facebook's products?); 2) technical analysis (lots of SQL); 3) analytical (questions about AB testing and the properties of the binomial distribution); 4) applied data (general reasoning about data problems); 3) A follow-up SQL interview Throughout the interview process, I realized that this was more of a business /product analyst role than a data science role, since the main focus was on SQL queries and stat101 probability questions (e.g. "what's the variance of the binomial distribution").