Machine Learning Researcher
Virtu Financial is a global financial services firm and market maker providing liquidity, execution services, analytics, and trading technology across multiple asset classes.
About Virtu Financial, Inc.
Virtu Financial, Inc. leverages market-structure expertise and technology to provide liquidity and transparent trading solutions across equities, ETFs, foreign exchange, futures, fixed income, cryptocurrency, and commodities. Its offerings include market making, execution services, liquidity sourcing, multi-asset analytics, data products, compliance tools, and broker-neutral workflow technology.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
Pursue high-impact machine learning research in a results-oriented environment, owning the modeling ecosystem from feature engineering through execution strategy and partnering with quantitative traders, researchers, and developers to build predictive models for financial markets.
Requirements
- Minimum 2 years of applied experience developing deep learning solutions.
- Experience applying machine learning methodologies across problem domains.
- Strong production mindset focused on tangible business outcomes.
- Proficiency in Python and contemporary deep learning frameworks.
- Advanced programming expertise in core PyTorch or JAX framework development.
- Comfort working cross-functionally with researchers, developers, and traders.
Responsibilities
- Investigate, evaluate, and prototype innovative machine learning and deep learning solutions.
- Develop deep learning models that directly impact P&L.
- Implement ML approaches for forecasting, feature engineering, and optimization.
- Conduct empirical ML research and iterate novel architectures in Python, PyTorch, and TensorFlow.
- Translate research methods across application areas using logical and mathematical reasoning.
- Partner with quantitative traders, researchers, and developers to transform market insights into predictive models.