Senior Principal Machine Learning Engineer (Fulfilment)

As Senior Principal Machine Learning Engineer (Fulfilment) at Grab, you will be the senior individual contributor lead architecting Grab's next generation of autonomous marketplace decisioning systems.

Working at the intersection of reinforcement learning, latent world models, and generative AI, you will guide technical strategy and translate research into production systems serving millions of users.

You will be the technical lead for a specialised team of scientists and engineers, reporting directly to the Head of Data Science.

Your responsibilities include:

• Architecting and implementing end-to-end reinforcement learning pipelines to jointly automate pricing, dispatch, and supply repositioning decisions.
• Designing latent world models and spatio-temporal marketplace state representations that capture complex supply-demand interactions and multi-temporal levers.
• Establishing formal off-policy evaluation frameworks, conservative value estimation baselines, and safety guardrails to de-risk live deployment in production.
• Building behavioural simulation systems using fine-tuned open-source large language models to forecast and simulate user and partner decision dynamics.
• Partnering with backend, streaming, and platform engineering teams to deploy low-latency, closed-loop decision agents into high-throughput production environments.
• Leveraging modern Generative AI development tools and AI-assisted workflows to accelerate prototyping, code quality, testing, and technical documentation across the team.

To qualify, you need:

• A PhD in Computer Science, Operations Research, Applied Mathematics, or a related quantitative discipline with at least 10 years of post-degree experience leading complex machine learning research and production initiatives (or Master’s degree with 12+ years of equivalent experience).
• Proficiency in reinforcement learning foundations, including Markov Decision Processes, stochastic control, offline RL, and reward shaping for sequential multi-objective decision systems.
• Proven experience deploying production ML/RL systems operating under real-time constraints, non-stationary data, and distribution shifts.
• Hands-on expertise with large language model adaptation and alignment, including supervised fine-tuning, parameter-efficient methods, and preference tuning.
• Hands-on proficiency in Python, PyTorch, and distributed training/compute platforms such as Ray, DeepSpeed, FSDP, or Megatron-LM.
• Experience engaging with AI tools and emerging technologies to enhance productivity, improve workflows, and contribute new ideas.

Grab offers various benefits, including term life insurance, comprehensive medical insurance, GrabFlex, parental and birthday leave, and a confidential Grabber Assistance Programme.