What the C++ track tests

Akuna's C++ interviews go beneath the language surface. Expect questions like what a cache line is and how false sharing degrades a multithreaded counter, when move semantics actually elide copies, the cost of virtual dispatch in a hot loop, and how you would build a single-producer single-consumer queue without locks. The trading path processes market data at microsecond scale, and interviewers select for engineers who think in memory layout and mechanical sympathy by default.

Design questions follow the same theme, a feed handler that parses millions of messages per second, an order gateway that must never send a duplicate, or diagnosing tail latency in a pinned-thread process. General distributed-systems answers miss the point here, the interesting problems are single-machine, kernel-bypass adjacent, and measured in nanoseconds.

The Python track and the broader loop

Python developers at Akuna build the research and data backbone, backtesting platforms, market data pipelines, and internal tools quants depend on. Interviews test Python at internals depth, iterator protocol, the GIL's real implications, memory behavior of dataframes, plus practical design like a pipeline that reprocesses a day of options data quickly and correctly. The bar is production engineering, not scripting.

Across tracks, Akuna teaches trading knowledge internally and its interview reflects that, a curiosity check rather than a finance exam. The firm is younger-skewing and known for a social, flat culture unusual among market makers, and final-round fit conversations genuinely matter, they are assessing whether you will thrive on a desk where developers, quants, and traders sit together.

Recommended Resources