Lead Machine Learning Engineer (Foundation Models)
Core
Architect and deploy proprietary foundation models and generative recommendation systems for Grab's Southeast Asia superapp marketplace.
Role type
Lead Machine Learning Engineer (Foundation Models)
Builds
Proprietary foundation models, generative recommenders, and large-scale distributed training infrastructure.
Domain
Southeast Asia marketplace / Generative AI / Foundation Models
Deliverable
production ML models
Required skills
Deep learning, transformer architectures, pre-training large language models, distributed training frameworks (FSDP, DeepSpeed, Megatron-LM), LLM post-training (SFT, LoRA, RLHF), Python/C++ software engineering, MLOps/LLMOps
Preferred skills
Generative retrieval, semantic tokenization, unifying retrieval and ranking with preference-aligned models
Technologies
PyTorch, Ray, GPU clusters, FSDP, DeepSpeed, Megatron-LM, LoRA, RLHF, DPO
Responsibilities
Design and orchestrate scalable pre-training pipelines for foundation models; Develop and operate large-scale distributed training frameworks; Design and deploy generative recommendation systems; Optimize transformer architectures (MoE, long-context attention); Lead code reviews and mentor junior engineers; Integrate custom models into live production environments
Seniority
Senior, hands-on IC with mentorship responsibilities