Machine Learning Researcher
Citadel SecuritiesAbout the Role
At Citadel Securities, Machine Learning Researchers on our Options team come from a wide range of industries and turn cutting-edge ideas and petabyte-scale data into bleeding edge models with direct trading impact. Our team of researchers iterate quickly, own decisions end-to-end, and operate with substantial autonomy, resources, and scope in a flat, no-bureaucracy environment.
- Own the full research lifecycle, from hypothesis, experiment design, model validation, risk/overfit controls, to deployment
- Conduct cutting-edge research and development in machine learning (e.g. LLMs) at scale with a focus on industry leading techniques and their applications in quantitative finance
- Ship models to production that move P&L in options markets—measured by clear, testable outcomes
- Prototype, test and iterate fast, taking great ideas from concept to trading in a very short space of time
- Discover alpha in high-dimensional data with deep learning, time-series, and representation learning
- Engineer scalable research pipelines from feature generation to distributed training and backtesting
- Develop trading intuition to translate insights into executable strategies
- Leverage large scale compute and data (petabytes; large budgets) to run ambitious experiments and push the frontier
Responsibilities
- Own the full research lifecycle, from hypothesis, experiment design, model validation, risk/overfit controls, to deployment
- Conduct cutting-edge research and development in machine learning (e.g. LLMs) at scale with a focus on industry leading techniques and their applications in quantitative finance
- Ship models to production that move P&L in options markets—measured by clear, testable outcomes
- Prototype → test → iterate fast The resources and support to take great ideas from concept to trading in a very short space of time
- Discover alpha in high-dimensional data with deep learning, time-series, and representation learning
- Engineer scalable research pipelines from feature generation to distributed training and backtesting
- Develop trading intuition to translate insights into executable strategies
- Leverage large scale compute and data (petabytes; large budgets) to run ambitious experiments and push the frontier
Requirements
- A curiosity to learn about financial markets, and excitement to understand microstructure, options dynamics, and volatility regimes on the job
- Masters or PhD degree in mathematics, statistics, physics, computer science, or another highly quantitative field, with advanced training and a strong research track record working on machine learning problems
- Deep knowledge of cutting edge large scale models and their training and design
- Training techniques (pre-training, fine-tuning, RL, RLHF), and optimization methods
- A results-oriented track record of having taken ML ideas from theory to measurable impact
- Strong math fundamentals (linear algebra, probability, optimization) and mastery of regression/ML for large scale data
- Hands-on with modern machine learning (sequence models/transformers, representation learning, regularization, cross-validation, causal/robust inference) applied in practice
- Bias to action & problem-solving demonstrated ability and comfort around owning decisions, iterating quickly, and simplifying complex problems to impactful solutions
- Curiosity about markets and enthusiasm to learn microstructure, options dynamics, and volatility regimes on the job
- Fluency in Python (NumPy, PyTorch) and the ability to write clean, modular, performant code for large-scale experiments