Quantitative Researcher
Wunder Fund has been involved in HFT since 2014. We trade on a bunch of exchanges worldwide, from traditional assets to crypto, with a daily turnover of > $10,000,000,000.
We are looking for strong hands for our Octopus team, our elite R&D unit consisting of a lead, a quant researcher, and an engineer. Both researchers are IPhO gold medalists with a strong background in ML, and the engineer is a skilled technician with extensive expertise in optimization and acceleration.
In less than a year, the team has developed several strategies from scratch that trade on non-obvious exchange signals. The secret to success lies in the combination of mathematics, logic, predictive power of ML, as well as the ability to iterate quickly and find non-trivial patterns where others do not see them.
What you will do
- Look closely at the data and find anomalies that are not visible to the naked eye (a challenge with a star, isn't it?).
- Develop trading strategies that make money on our exchange simulator — a backtest.
- Improve the backtest, making it even closer to reality (so that the strategies developed in the previous point stop working so well).
- Analyze the performance of the best strategies in live trading, look for discrepancies with the backtest, and repeat the previous steps until perfect convergence.
Tech Stack
- For experiments and prototyping: Python, Polars/pandas, scikit-learn/Catboost/PyTorch.
- For ultra-fast backtesting and computationally intensive calculations: Cython / C++ / Rust.
- For quick prototypes and research iterations: Codex / Cursor.
Ideal Match
This role has narrow requirements — we know exactly who will strengthen the team and achieve success. For a mutual match, you should have something from the following:
- Experience in an AI startup: you have worked in an idea-driven environment where success is determined not by the number of people, but by their quality and desire to do something cool, where you need to generate ideas yourself and be passionate about their implementation.
- Experience in R&D projects in scientific laboratories (such as Yandex Research, AIRI Institute, ISPRAS, etc.). Your project moved science forward, you fought with uncertainty in cutting-edge areas of ML/mathematics/optimization/etc.
- Successes in Kaggle: the title of Kaggle Competition Expert or higher, as well as solo gold or successful participation in similar competitions with tabular data.
Also required
- Skill of intense data scrutiny: you can immediately share several interesting stories about how careful examination of raw data helped find an non-obvious anomaly / gain insights into the world's workings and solve a problem.
- Experience in rapid experimentation. You will need to generate hypotheses, test them, discard non-working ones without regret, and search for new ones. It is important not to cling to an idea, but to thoughtfully try different approaches, learning from failures.
- Experience using AI Coding Assistants to accelerate R&D cycles (Codex / Cursor / alternative dev environment).
- This role does NOT require: experience in HFT and trading, expertise in finance or economics.
Why these requirements
The work of a researcher in HFT is work in conditions of total uncertainty, where most hypotheses turn out to be unsuccessful. Therefore, it is very important to be able to move from failure to failure without losing enthusiasm, iterate quickly, try unconventional approaches, and not give up. If you are driven by the search for one brilliant idea among thousands of unsuccessful ones — you have found your place.
Conditions
- Minimum bureaucracy, a small team of strong specialists, always someone to discuss ideas with.
- Full remote work, we compensate for co-working spaces. After the probationary period, we assist with relocation / obtaining a residence permit — if desired. If you are of draft age and in the Russian Federation — we will help with deferment.
- Workation adventures. Twice a year, we rent a large villa for a couple of months where you can work, live, and hang out together (we've already been to Thailand, Turkey, Bali, and Georgia).
