Deep Learning Quantitative Researcher

About the Company and Role

Join Millennium in Singapore as a Deep Learning Quantitative Researcher, contributing to cutting-edge quantitative alpha research by implementing advanced deep learning techniques.

Primary Responsibilities

• Architect and develop the firm’s central deep learning systems, covering all stages from data processing and distributed training to evaluation and deployment in production environments.
• Lead substantial research initiatives by applying deep learning methods, managing the entire empirical workflow including problem definition, model conception, training, validation, and attribution of results.
• Maintain strict research standards within a challenging signal-to-noise context by enforcing out-of-sample testing rigor, preventing data leakage, and benchmarking against simpler baseline models.
• Serve as the deep learning specialist within the company by guiding architecture choices, analyzing training processes, reviewing model designs, and setting evaluation and deployment criteria.
• Enhance cross-team integration of model training and inference through standardized interfaces and reusable components.

Candidate Qualifications and Experience

• Possess a PhD in Computer Science, Engineering, Physics, Mathematics, or Statistics with 3–5 years of relevant professional experience applying deep learning to large-scale problems, preferably in quantitative finance. Exceptional PhD research plus experience at leading AI/technology firms is also valued.
• Demonstrated ownership of the full lifecycle of deep learning models in either production-grade systems or published research projects.
• Expertise in Python programming and modern deep learning frameworks.
• Hands-on experience with large-scale distributed or multi-GPU model training, including skills in mixed precision and performance profiling for optimization.
• Strong foundational knowledge in statistics, optimization techniques, and machine learning theory.

Technical Skills and Knowledge

• Proficiency in contemporary deep learning architectures with the discernment to prefer simpler models when warranted.
• Experience handling low signal-to-noise learning environments via regularization, ensembling, and robust out-of-sample validation protocols.
• Expertise managing large datasets, including use of efficient columnar data formats, streaming data loaders, and constructing datasets with point-in-time correctness.
• Familiarity with tools for managing experiments such as tracking systems, hyperparameter tuning, and environments ensuring reproducible research.
• Additional skills in C++ programming or CUDA optimization are advantageous. Knowledge of large language model (LLM) tools to accelerate research is a plus.

Professional Traits

• Exemplary research judgment, capable of designing precise experiments and decisively discarding ideas based on evidence.
• Proactive and collaborative approach, fostering strong working relationships between research and engineering teams.
• Commitment to high ethical standards, especially when managing sensitive data and models.
• Growth-oriented mindset, continuously updating skills and adopting effective new techniques.
• Exceptional communication skills, able to explain model behavior and uncertainties effectively to both technical and non-technical stakeholders.

Additional Information

Interested candidates should reference REQ-30088 in their application emails to QuantTalentEUR@mlp.com.