Senior Data Scientist - Computer Vision & Multimodal LLMs

About the role

The Integrity team protects users and businesses across fraud, risk, digital identity and safety, using data science to build AI solutions. This Senior Data Scientist role focuses on computer vision and multimodal large language models after training, reporting to the Senior Data Science Manager and working from the Grab Singapore One North office. The work supports user onboarding, verification workflows and downstream applications.

Key responsibilities

• Design, train and fine-tune state-of-the-art models, including OCR and Vision LLMs, to extract information from complex documents and scale KYC and onboarding operations.
• Architect and own end-to-end document processing pipelines, from initial concept to live production models, covering identity verification.
• Deploy models into a high-throughput production environment, set up monitoring for production metrics and iterate to improve performance, latency and efficiency.
• Research and prototype with the latest AI advances, turning academic research into product impact.
• Explain technical concepts, model trade-offs and experimental findings to stakeholders in product, policy and operations.

Required profile

• Degree in Computer Science, Electrical or Computer Engineering, Mathematics or Statistics, or a related technical discipline.
• At least 3 years of experience in tech companies or academic institutions in a similar role.

Required skills

• Proficiency in Python, Spark and SQL.
• Familiarity with version control systems such as git and git-flow, and an understanding of the full software development life cycle.
• Experience with deep learning frameworks such as TensorFlow or PyTorch.

What we offer

• Term life insurance and comprehensive medical insurance.
• GrabFlex benefits package tailored to individual needs.
• Parental and birthday leave, plus volunteering leave.
• A confidential assistance programme for Grabbers and their families.
• FlexWork arrangements such as differentiated hours.