Senior Data Scientist (Trust & Fraud)

Data Science; Selangor; On-site; MYR 90,000 - 140,000; Full time

Job summary

Data Science in Malaysia is seeking a talented Data Scientist to combat fraud at scale. You will analyse transactional data, build and deploy ML models, and work with cross-functional teams to integrate fraud detection systems on-site in Petaling Jaya.

You will push the boundaries of ML and LLM applications, collaborating with software and product engineers, and leveraging GNNs, Transformers, and open-source LLMs to accelerate development and improve security across the Grab ecosystem.

Qualifications

• Partner with teams to architect scalable data science solutions that translate complex operational challenges into strategic wins, integrating both traditional ML and agentic LLM systems.
• Conduct cutting-edge research to incorporate the latest advancements in algorithms and generative AI into our defense systems to counter emerging fraud tactics.
• Master data orchestration by preparing, augmenting, and combining diverse data types to build high-fidelity training datasets, including using LLMs to generate synthetic data and labels.
• Design, train, and fine-tune model architectures, utilizing Graph Neural Networks (GNNs), Transformers, fine-tuned open-source LLMs (like Qwen), and Gradient Boosted Trees.
• Lead the end-to-end lifecycle of your models—from production deployment to performance monitoring—iterating alongside software and product engineering teams.
• Leverage Generative AI tools (like coding assistants and analytical co-pilots) in your daily workflows to accelerate code generation, automate testing, and enhance your overall productivity.

Responsibilities

• Partner with teams to architect scalable data science solutions that translate complex operational challenges into strategic wins, integrating both traditional ML and agentic LLM systems.
• Conduct cutting-edge research to incorporate the latest advancements in algorithms and generative AI into our defense systems to actively counter emerging fraud tactics.
• Master data orchestration by preparing, augmenting, and combining diverse data types to build high-fidelity training datasets, including using LLMs to generate synthetic data and labels.
• Design, train, and fine-tune model architectures, utilizing a toolkit that includes Graph Neural Networks (GNNs), Transformers, fine-tuned open-source LLMs (like Qwen), and Gradient Boosted Trees.
• Lead the end-to-end lifecycle of your models—from production deployment to performance monitoring—iterating alongside software and product engineering teams.
• Leverage Generative AI tools (like coding assistants and analytical co-pilots) in your daily workflows to accelerate code generation, automate testing, and enhance your overall productivity.

Job description

Get to Know Our Team

We are the architects of trust. Our mission is to shield the Grab ecosystem from evolving fraud and safety threats by turning massive datasets into applicable intelligence. From deploying sequence-based models for payment risk to applying graph algorithms that unmask complex money laundering networks, we operate at the intersection of deep learning and platform security. We don't just react; we innovate, researching the latest methods to neutralise tactics before they even surface.

Get to Know the Role

You will fight fraud by analysing transactional data, developing and deploying machine learning models, and collaborating with cross-functional teams to ensure the seamless integration of fraud detection systems. Reporting to the Data Science Manager II, you will work as an Individual Contributor in a full-time, on-site role at our office in Petaling Jaya, Malaysia. In this role, you will push the boundaries of machine learning and LLM applications at a massive scale, growing your expertise in cutting-edge agentic systems.

The Critical Tasks You Will Perform

• You will partner with teams to architect scalable data science solutions that translate complex operational challenges into strategic wins, integrating both traditional ML and agentic LLM systems.
• You will conduct cutting-edge research to incorporate the latest advancements in algorithms and generative AI into our defense systems to actively counter emerging fraud tactics.
• You will master data orchestration by preparing, augmenting, and combining diverse data types to build high-fidelity training datasets, including using LLMs to generate synthetic data and labels.
• You will design, train, and fine-tune model architectures, utilizing a toolkit that includes Graph Neural Networks (GNNs), Transformers, fine-tuned open-source LLMs (like Qwen), and Gradient Boosted Trees.
• You will lead the end-to-end lifecycle of your models—from production deployment to performance monitoring—iterating alongside software and product engineering teams.
• You will leverage Generative AI tools (like coding assistants and analytical co-pilots) in your daily workflows to accelerate code generation, automate testing, and enhance your overall productivity.