ml engineer for digital assets
Описание:
Wintermute is an algorithmic trading company specialising in digital assets. It provides liquidity across cryptocurrency exchanges and trading platforms, OTC trading solutions, and support for blockchain projects and traditional financial institutions entering crypto. Its ventures arm invests in early-stage DeFi projects.
Задачи:
Develop ML-based alpha generation models using high-frequency order book and market microstructure data Design and maintain data pipelines, preprocessing, and feature extraction workflows for streaming tick data Research and implement deep learning architectures for short-horizon forecasting and signal extraction Collaborate with quant researchers and developers to integrate models into live trading environments Optimise inference latency and robustness, ensuring models behave safely under live market conditions Refine model quality through systematic backtesting, live evaluation, and monitoring
Требования:
Degree in Computer Science, Machine Learning, Applied Mathematics, or a similar quantitative discipline Strong Python programming skills and familiarity with ML libraries Proven track record applying ML/DL to real-world problems Familiarity with time-series modeling, signal extraction, or high-frequency data Experience developing ML infrastructure, including data pipelines, experiment tracking, and versioning Будет плюсом: Experience in finance, trading, or quantitative research; publications, competition results such as Kaggle or academic ML contests, or open-source contributions; familiarity with C++, CUDA, or low-latency systems
Условия:
Performance-based compensation with significant earning potential Pension and private health insurance Team meals, festive celebrations, gaming events, and company-wide team-building events Central London office with table tennis, foosball, personalised desk configurations, and a team breakout area UK work permits and relocation assistance are available Significant flexibility in working from home and working hours; fully remote candidates are not accepted
#J-18808-Ljbffr