Capital One Data Scientist Interview Experience — Two Attempts, a Cancelled Headcount, and an Energy Case Interview

My Capital One interview process has been quite a rollercoaster so far — over the past month or so I've interviewed twice.

Timeline

The first time was through a referral from a friend, for a newly posted analytics role:

• 1/16: HR Call scheduled
• 1/24: HR Call rescheduled to 1/31 — the reason was that when the original interview time came, I sat on my phone waiting for over half an hour, then HR finally showed up with an email saying there'd been a fire drill so it had to be rescheduled
• 1/31: HR Call — since it was a Friday, HR told me at the end that they'd send the immigration assessment the following week
• 2/6: System rejection email — I followed up with HR to ask why, and HR replied that the position had been cancelled, they no longer needed to hire for that headcount

The second time, HR reached out to me directly on LinkedIn — this HR was great, and things moved fast:

• 2/11: HR reached out
• 2/13: HR had me submit an application on the official site + scheduled an HR call
• 2/20: HR Call
• 2/20: received the immigration questionnaire
• 2/21: received an HR email confirming the immigration questionnaire submission
• 2/24: mini case interview scheduled
• 2/26: Capital One workshop for the case interview
• 3/5: mini case interview

Mini Case

I got the "energy" case question, which is one that's come up a lot on the forum. The whole interview lasted an hour.

The interviewer started with a self-introduction, then walked me through what the case interview would involve.

For the long background intro at the start, honestly I didn't fully catch it in English — I only picked up two key pieces of information: what the company does (an energy company), and what problem we're solving (we want to invest in new energy):

• Question 1: what factors do you consider?
• Question 2: calculation — figure out how much energy needs to be produced to hit a 10% annual return on investment. I calculated 6.25M.
• Question 3: now there's a constraint limiting production to 5M — what can you do to still hit the 10% annual ROI?
• Question 4: there are two options — one is solar, the other kind of energy I couldn't quite make out — asking how much production is needed to break even. I calculated 2.5 years for both. I actually got the first option wrong at first — I assumed the $30 variable cost applied to both energy types, but the interviewer pointed out that solar has no variable costs.
• Question 5: which option would you choose
• Question 6: at the end, you get 30 seconds to give a summary report — I hadn't prepared for this one, so it honestly came out pretty stiff.

With about 5 minutes left in the interview, the interviewer asked how I felt about the process, then had me ask him a question about Capital One. I told him the whole thing felt really packed, and he replied that "power day" would feel even more packed than this — which made me wonder if that was some kind of hint, haha.

Other thoughts:

Interview experience: the interviewer I got was super nice!! Really patient — he confirmed every single question and even every single data point with me! When he was reading out the question, honestly I didn't understand anything... The first time he read out the numbers, I could tell he was reading slowly but I still didn't retain any of it. He really did go back over the numbers with me again and again to confirm and correct them, and guided me toward the right answer! And of course, he was also really good-looking!!

Looking back at the whole interview, I felt like what they're testing isn't whether you get the correct answer, but your thought process and how you demonstrate that thought process to the interviewer. For example, on question 4 I got it wrong at first, but when he pointed it out he said it was fine, the result didn't matter — he cared more about how I arrived at the number. So I think the most important things in this interview are logical thinking and communication.

As for prep, I think there are two approaches: one is to grind through past interview write-ups on the forum, since the chance of hitting a repeat question is pretty high; the other is to study case interview frameworks. Since I had limited prep time and also like to slack off, I only studied case interview frameworks online and summarized the common case question types with their corresponding frameworks — for forum write-ups, I only went through the two most recent case reports.