Two rounds at Bloomberg, and a language barrier

Interviews   · 3 min read

CompanyBloomberg
TeamEngineering, entry-level
Role2026 Software Engineer
LocationNew York. Both rounds virtual, on Zoom and HackerRank
DateDecember 2025 to January 2026
How it startedCold application
Rounds2
OutcomeRejected, two days after round 2

This is the one where I reached the optimal solution and still lost. It is also the one where I wrote back after the rejection.

The loop

RoundWhat it wasProblemSelf Rating
1DSA, one interviewer, 60 minutesGraph. MediumHire
2DSA, two interviewers, 60 minutesGraph, shortest path. Solved optimally in C++. A second problem was mentioned and never reachedHire

Round 1 was in December. I cleared it, and a month later I was booked into the second.

The format told me before the email did

The confirmation for round 2 said there would be a minimum of one technical interview, and that depending on interviewer availability and on performance, other people might be invited to join for more. It asked me to keep three or four hours free.

I finished the first interview and emailed the recruiter to ask whether more rounds were coming. The reply was that there would be no more that day. In a format where further interviews depend on how the first one goes, that was the answer, two days before the rejection arrived.

What happened in round 2

I wrote the solution in C++, which is the language I think fastest in, and which Bloomberg’s own instructions said I was free to choose. It became clear early that neither interviewer was comfortable reading it. One said outright that they had limited familiarity with the language.

So the hour went to explaining the implementation rather than evaluating the solution: what a construct did, why a line was there, what a standard library call returned. I reached the optimal approach. A second problem had been mentioned at the start, and with time still on the clock it never came up.

What I sent afterwards

The rejection came two days later. I wrote back to the recruiter. The core of it:

The interview placed strong emphasis on problem-solving, which I appreciated and was able to address by arriving at the most optimal solution during the discussion. […] A significant portion of the interview time was spent explaining the implementation rather than evaluating the solution itself.

I’m sharing this feedback because I believe interviewer-language alignment is important for fair technical evaluation when problem-solving is a primary signal.

There was no reply. That makes two loops on this site that ended in silence, though unlike Amazon, this one at least sent the rejection first.

The part that is mine

It would be easy to file this entirely under a bad draw of interviewers, and I do think the evaluation was weaker for it. But there is a version of this I could have prevented.

Bloomberg said I could choose my language. It did not say the people on the call would read it fluently, and I did not ask. Thirty seconds at the start, asking whether C++ worked for them and offering to switch to Python, would have cost nothing.

In an interview the language is not only how you implement the answer. It is part of how you communicate it, and I had been treating it as only the first.

A mountain silhouette at dusk, with the words: Success is not final, failure is not fatal. It is the courage to continue that counts.
Image from motivation_mondays on Instagram. The quote is usually credited to Winston Churchill, who never said it. The closest early version is a 1938 Budweiser advertisement.

← All writing