Hudson River Trading (HRT)

Official guidance

Hudson River Trading is an algorithmic trading firm. Its October 2025 engineering post splits engineering into Trading Tech, focused on latency (exchange market data, order entry, FPGA/ASIC acceleration, trade processing), and Research & Development, focused on scale (storage, data centers, CPU/GPU clusters, job scheduling, data pipelines and research tools), each with platform teams and engineers embedded with trading teams. Work ranges from Verilog and low-level C++ to Python services and TypeScript/Python full-stack tools.

Relevant roles: General software engineering, Low-level software systems, Backend engineering, Distributed systems engineering, Infrastructure / platform engineering, Production engineering / SRE, Frontend / full-stack engineering. Domains: Algorithms, OS & concurrency, Networks, Distributed & cloud, Software engineering, Interview skills.

What the sources establish

HRT’s October 2025 HRTbeat post, written by a long-tenured member of its engineering organization and limited to engineering roles, describes the current process: a series of Zoom calls (exploratory conversations or technical phone screens), technical interviews for every candidate, then onsite interviews (HRT prefers at least some in person) generally made up of coding and debugging rounds, technical design discussions and team fit. Candidates choose C++, Python, or TypeScript/Python for full-stack roles, may be asked to analyze code in an unfamiliar language, and take the same technical tests regardless of seniority. The post lists the most common failure reasons and calls using LLMs to get through early stages cheating. Not established: the number and length of interviews, online assessments or take-homes, and detailed tool rules.

Answers to Questions I Often Get: Engineering and Interviewing at HRT
Official
Hudson River Trading (HRTbeat). Published 2025-10-02 (updated 2025-10-03); read 2026-10-05. Read in full for this page.
Scope: Engineering roles only, at all levels; the post explicitly excludes Algo Development (quantitative research and trading) and other job families.
Documented by this source
- The process starts with a series of Zoom calls that, depending on role and candidate, are exploratory conversations or technical phone screens; all candidates eventually do technical interviews.
- Interview language: C++, Python, or TypeScript/Python for full-stack roles; some problems may ask you to analyze code in an unfamiliar language, graded with your familiarity in mind.
- Interviewing in C++ implies a fair amount of low-level knowledge of how computers work; interviewing in Python still implies some systems knowledge, such as how virtual memory works.
- Onsite interviews, preferably at least partly in person, generally consist of coding and debugging rounds, technical design discussions and team fit, varying slightly by role.
- HRT avoids “burst of insight” puzzle questions and aims to test strong fundamentals rather than obscure tricks.
- The same technical tests are given at all levels, and candidates should expect to program during the interview whatever their seniority.
- Most common failure reasons, roughly in order: shallow understanding of fundamentals, communication issues, not listening or understanding the problem, avoiding direct practical solutions, arrogance, indifference, and cheating.
- Most engineering roles share one pipeline, so team placement can be adjusted later in the process.
Outside or AI tools, according to this source: Calls using LLMs and similar tools to get through earlier stages cheating, says later stages make cheating much harder, and says cheating candidates are not welcome to interview again. No other tool rules are given. Follow the instructions in your own assessment invitation.
- A personal post on HRT’s official blog rather than a formal candidate guide.
- Gives no interview counts or durations and does not mention online assessments or take-homes.
- Automated downloads of the article page were blocked (HTTP 403); the full text was read through HRT’s own site API and cross-checked with a fetch tool.
Open source(opens in a new tab)Not checkable automatically, 5 Oct 2026
Check note: The site refused or challenged the automated request (HTTP 403). This is common for bot protection and does not mean the link is broken.

Our preparation recommendation

This is our suggestion based on the role and the sources above. It is not a statement by Hudson River Trading (HRT).

Our preparation suggestion, not an HRT statement. HRT says its onsites combine coding, debugging and technical design, that it tests fundamentals rather than puzzle tricks, and that the depth expected depends on your interview language: low-level knowledge for C++, and systems basics such as virtual memory even for Python. We therefore weight language internals, memory and OS fundamentals, debugging and practical, direct coding, plus design for latency-bound or scale-bound systems that match the two engineering divisions the post describes. Do not rely on AI tools at any stage; HRT treats that as cheating.

C++ fundamentals: lifetime, ownership and undefined behaviorNot started
If you interview in C++, HRT expects low-level understanding; its illustration of shallow fundamentals asks whether you know how most compilers implement virtual classes.
Python fundamentals for interviews and research codeNot started
If you interview in Python, know the internals of what you use; HRT’s illustration asks whether you know how a Python list’s memory grows.
Virtual memory and memory managementNot started
Named by HRT as systems knowledge expected even of Python interviewees.
CPU caches, locality, coherence and NUMANot started
Inference for Trading Tech roles, whose main technical challenge the post names as latency.
Profiling, bottlenecks and benchmarking pitfallsNot started
Inference: latency and scale problems are diagnosed by measurement, which helps in design and debugging discussions.
Debugging, code reading, review and refactoringNot started
Debugging rounds and reading code in an unfamiliar language are documented parts of the process.
Coding fluency and standard-library costsNot started
Coding rounds reward simply coding a practical answer in your chosen language.
Correctness, invariants, edge cases and testingNot started
Careful, working code is the opposite of the “avoiding real work” pitfall HRT describes.
System design methodNot started
Technical design discussions are part of the onsite.
Explaining solutions aloudNot started
Communication and listening are among the most common failure reasons HRT lists.

Confirm with your recruiter

Which Zoom calls you will have (exploratory conversation or technical screen) and how many onsite rounds of each kind: coding, debugging, design and team fit.
Your level and target area (Trading Tech or Research & Development), since roles share a pipeline and placement can shift.
Your interview language (C++, Python, or TypeScript/Python for full-stack) and how deep the low-level questions go for it.
Tool rules for each stage, including IDE features, documentation and AI assistants; the post treats LLM use in early stages as cheating.
Whether there is any online assessment or take-home (the post mentions none).
Which onsite interviews will be in person, plus travel logistics.
Whether debugging rounds use a provided codebase, and in which language.

Limitations of this page

One blog post is the only source; it gives no counts or durations.
Scoped to engineering; Algo Development (quantitative research and trading) interviews are not covered.
The latency and performance emphasis is our inference from the post’s description of the engineering divisions.