Boston Consulting Group AI Engineer interview questions

The BCG X AI Engineer hiring pipeline includes an automated interview, an online test, technical system design interviews, a unique business case study, and partner fit rounds. x x x

1. Codesignal GCA - There is a lot of information online about this assesment. It consists of 4 questions, similar to leetcode: 2 easy, 1 medium (most probably going to have to deal with a matrix and lists, being the most lines of code among all the questions) and 1 hard (where you have to optimize the speed and memory usage, possibly dynamic programming. 2. Live coding interview - It will involve at least one algorithm, dealing with data manipulation, but wouldn't consider to be harder than any of the GCA questions. And a SQL data extraction question, requiring table joins and group by aggregation. Honestly, I felt it easier, but you do have to explain your thinking thoroughly. Don't keep quite, even when making a mistake; talk to the interviewer as he could lead you on the correct path. This is more about knowing what the code is doing, rather than memorizing certain algorithms. 3. 2 technical case interviews - Both of these had me going over designing a digital system to serve a certain purpose. You don't have to know any specific framework, except a more general view (using SQL vs NoSQL, serverless computing advantages and disadvantages, generic API request/response, latency, when to use RAG in AI and how to make use of data to solve a problem). It is not about knowing every single detail in the technologies you propose, but understanding their purpose, drawbacks and how to deal with them. 4. Behavioral interview - This is a more relaxed scenario in which you talk a little about yourself, your experience and your expectations of the job. They may mention some technical concepts for you to explain, but nothing overly complicated. Make sure to show some enthusiasm and that you are an organized person wanting to learn from the job and colleagues.