Senior Multimodal ML Research Lead

cambridge centre for advanced research and education in singapore ltd.; Singapore; On-site; SGD 180,000 - 280,000; Full time

Job summary

Cambridge CARES, the University of Cambridge’s research centre in Singapore, is seeking an experienced ML researcher to lead the BloodCounts! project.

You will design and train a multi-modal foundation model using anonymised clinical data across imaging, tabular and signal modalities, coordinating with international partners. You will drive interpretability, uncertainty quantification and robust deployment strategies, while mentoring a team of ML researchers and contributing to open-source

Qualifications

• Substantial machine learning research experience in an academic, data-intensive or clinical context, typically 5–10 years postdoctoral or equivalent.
• Strong track record in modern deep learning, with depth in generative modelling.
• Experience developing foundation models or large-scale self-supervised models and adapting them to downstream tasks.
• Proficiency in Python and PyTorch, including distributed and multi-GPU training.
• Experience with integrating diverse modalities within a unified architecture.
• Understanding clinical data modelling challenges including missing data and bias.
• Experience with interpretability, uncertainty quantification and deployment considerations.
• Knowledge of version control, experiment tracking, reproducible pipelines and data/model versioning.
• Ability to lead ML researchers and set technical direction.

Responsibilities

• Deliver all ML deliverables of the project, designing, training and evaluating the multi-modal foundation model.
• Lead development of representations across tabular, imaging and other modalities.
• Advance interpretability and uncertainty quantification for clinical biomarker discovery.
• Coordinate federated co-development across the BloodCounts! consortium for equitable performance.
• Set scientific and technical direction for the ML team and mentor researchers.
• Work with engineering to maintain reproducible ML tools and workflows.
• Contribute to manuscripts and open-source releases.
• Stay updated with rapid AI research developments.
• Collaborate with clinical and non-technical stakeholders.

Skills

Deep learning

Python

PyTorch

Multimodal ML

Federated learning

Model deployment

Experiment tracking

Academic research

Communication

Mentoring

Git

Distributed training

Reproducibility

Tools

MLflow

Weights & Biases

DVC

Job description

Cambridge CARES, the University of Cambridge’s research centre in Singapore, is seeking an experienced ML researcher to lead the BloodCounts! project.

You will design and train a multi-modal foundation model using anonymised clinical data across imaging, tabular and signal modalities, coordinating with international partners. You will drive interpretability, uncertainty quantification and robust deployment strategies, while mentoring a team of ML researchers and contributing to open-source