Hudson River Trading’s loop is almost entirely technical, and behavioral questions take only a small share of it. Fit questions come in recruiter conversations that TraderMath says can be interleaved with the technical rounds, in short segments of the onsite, and in at least one 2024 New York systems loop where behavioral questions were one of five onsite discussions. The questions asked most are why HRT, what you want to work on, a performance or debugging story from your own work, and how you work with other engineers and researchers.
Because the technical bar decides most outcomes, keep these answers short and concrete. Round details are on the Hudson River Trading interview questions page and candidate reports are on the Hudson River Trading interview experience page.
Where fit questions appear in the HRT loop
| Stage | What gets asked |
|---|---|
| Recruiter call | Background, why HRT, which track (software engineering or algorithm development), timeline |
| Phone screens | A short introduction before coding |
| Onsite | One behavioral discussion in some loops; project questions inside technical rounds |
| Between rounds | Recruiter check-ins on motivation and team interest |
Common Hudson River Trading behavioral questions
1. Tell me about yourself
Sample answer: “I’m a senior in computer engineering. I spend most of my time in C++: I wrote a lock-free queue for our robotics team’s sensor pipeline that cut end-to-end latency from 9 ms to 2 ms. I like work where you can measure the result in microseconds, which is why I’m interviewing with trading firms and HRT in particular.”
2. Why Hudson River Trading?
Sample answer: “HRT started in US equities and now trades on nearly every electronic market, and the code that does it is written in-house. I want to work where performance is the product. I also liked that your interviews go into how the machine works; that’s the part of computer science I enjoy most.”
3. Software engineering or algorithm development: why this track?
Framework: Say what you like doing for hours. Software engineers build the trading systems and infrastructure; algorithm developers research and build strategies with heavy probability and data work. Give one project that shows the match.
4. Tell me about a performance problem you solved
STAR sample: “Situation: A market-data parser in our trading club dropped messages at the open. Task: Handle peak rate without drops. Action: I profiled with perf, found 40% of time in heap allocation, switched to a preallocated ring buffer and parsed in place. Result: Throughput went from 300k to 2.1M messages per second with zero drops in a replay of the busiest day.”
5. Tell me about the hardest bug you’ve found
STAR sample: “Situation: Our simulator gave different P and L on two machines. Action: I narrowed it to one function by bisecting outputs, then found an uninitialized variable that UBSan flagged in seconds. Result: Results matched across machines, and I added sanitizer builds to CI.”
6. Describe a time you worked closely with someone outside engineering
Sample answer: “I built a backtesting tool for two economics students. They kept asking for features I thought were cosmetic until I watched them use it and saw they were re-running hour-long jobs to change one chart. I added result caching, and their iteration time dropped from an hour to two minutes.”
7. Tell me about a disagreement on a technical decision
STAR sample: “A teammate wanted std::map for an order book; I proposed a sorted vector. We benchmarked both on our real message mix: the vector was 3x faster for our book depth. He agreed, and I documented the depth at which the map would win.”
8. How do you make sure your code is correct?
Sample answer: “I write tests before submitting, including edge cases nobody gave me. On an OA with few provided tests, I’d write 5 to 10 of my own per problem. For anything concurrent I add a stress test and run it under ThreadSanitizer.”
9. What do you do when you’re stuck in an interview?
Sample answer: “I say out loud what I know and what I’m unsure of, try a small example, and propose a brute force first so we have something correct to improve on.”
10. What would you like to work on in your first six months?
Sample answer: “Something on the critical path, even a small piece, so I learn how HRT tests and deploys code that trades. Ideally a latency-sensitive component where I can measure my change.”
11. Do you have questions for us?
Strong questions:
- “How do engineers and algorithm developers split ownership of a strategy’s code?”
- “How do you test a change before it trades?”
- “What C++ standard does the main codebase use, and how do you decide when to move up?”
What HRT interviewers listen for in fit answers
- Measured results: microseconds, messages per second, memory used.
- Hands-on low-level work: profilers, sanitizers, debuggers, reading assembly.
- Self-testing habits: HRT’s OA gives few test cases, and interviewers notice who writes their own.
- Clear track choice: knowing whether you want to build systems or research strategies.
Answers that hurt at Hudson River Trading
- A “why HRT” answer about pay or prestige with no technical reason
- Performance claims with no measurement behind them
- Team stories where you can’t say what you personally wrote
- Unclear interest between software engineering and algorithm development
Similar companies
Citadel · Jane Street · Susquehanna (SIG) · D. E. Shaw · NVIDIA