Machine Learning Researcher

Belvedere Trading is a leading proprietary trading firm proudly headquartered in downtown Chicago, with a satellite office in Boulder, Colorado. Our teams work hard to provide liquidity to the market through their market-making activities by combining their deep expertise in commodities, interest rates, exchange-traded funds (ETFs), and equity index options with a cutting-edge technological system. Our trading models and software systems are continually developed, optimized, and upgraded in order to stay competitive in an ever-changing market. We work closely with our technology team to continuously and iteratively improve our proprietary solutions.
High-performance proprietary development is the source of our success and competitive advantage at Belvedere, further fueling our passion for performance. We are a team driven by intellectual curiosity, seeking answers that will change not only how we trade in this technological age, but also the future landscape of the trading industry. We place a high premium on defining, developing, and deploying high-performance trading software using a team-based, holistic development approach. We are looking for passionate team members whose contributions will be critical to our continued success.

This role is focused on developing and refining self-adaptive, feedback-driven models that can learn from and react to rapidly changing environments, potentially replacing or augmenting human decision-making. You'll be joining a collaborative team of researchers and engineers, working on cutting-edge ML applications in a high-impact domain. If you thrive in a fast-paced environment and have a passion for solving complex challenges, we'd love to hear from you.

What you'll do

• Research and develop machine learning models for real-time, closed-loop systems with feedback
• Embrace the uncertainty persistent in the marketplace to balance the risk-reward in each individual trade
• Design and test adaptive algorithms capable of evolving with incoming data and shifting regimes
• Build models that simulate or replace human decision-making in complex, uncertain systems
• Integrate control theory principles to improve model stability, responsiveness, and robustness
• Collaborate with engineers to deploy research into production environments with real-time constraints
• Continuously iterate and improve models based on system performance and market dynamics
• Communicate findings and methodologies clearly with other researchers, engineers, and stakeholders
• Develop best machine learning practices for larger quantitative group

What you'll need

• Ph.D. in Computer Science, Engineering, Applied Mathematics, or a related quantitative field
• 5+ years of experience applying machine learning in feedback-loop systems, robotics, autonomous decision-making, or control systems
• Familiarity in dealing with uncertainty in problem statements
• Strong knowledge of machine learning algorithms, especially in reinforcement learning, online learning, or adaptive control, with practical experience deploying ML models in production
• Solid programming skills in Python (preferred), and experience with ML libraries like PyTorch, TensorFlow, or JAX
• Strong analytical thinking and a track record of tackling open-ended, systems-level problems
• Demonstrated ability to work with Data Engineers to deliver robust data pipelines
• Financial market knowledge is a plus, but not required (we value technical depth and research excellence first and foremost)
• Active team player motivated by a fast-paced environment and conviction to focus on difficult, long-duration problems that require iteration
• Availability to work in a hybrid structure out of our Chicago or Boulder offices
  $140,000 - $240,000 a year Additionally, certain positions at Belvedere Trading are eligible for discretionary bonuses. Our employees have access to a variety of benefits, which can be found on our website here.