Senior Researcher, Multi-Modality

Tencent's Lightspeed Tech Center seeks a Senior Researcher, Multi-Modality to develop multimodal foundation models and agentic AI for in-game applications. The role connects research questions drawn from gaming scenarios with solutions that can be deployed across game products at scale. The position is full-time and onsite at Singapore-CapitaSky.

Responsibilities

• Pioneer research in multimodal understanding, post-training, reasoning, grounding and agent planning for gaming scenarios. Improve vision-language and multimodal large-model understanding across images, video, text and audio.
• Lead post-training work using supervised fine-tuning (SFT), RLHF, RLAIF, reward modeling and preference alignment to improve model capability and reliability.
• Design and run experiments from data preparation and benchmarking through training, evaluation and iteration. Work with engineers to move research prototypes into production for diverse in-game uses.
• Publish research at top-tier conferences and contribute to the research community.

Qualifications

• PhD in Computer Science, machine learning or a related field, or equivalent research experience.
• Expertise in at least one of multimodal understanding and vision-language reasoning; LLM or VLM pre- or post-training; reinforcement learning and LLM-based agents; or representation learning.
• Strong implementation skills in modern ML frameworks such as PyTorch or JAX, plus demonstrated ability to conduct complex experiments, analyze results and iterate quickly.
• Consistent first-author publications at leading venues such as NeurIPS, ICML, CVPR or ACL. Ability to turn real business problems into research and deploy solutions; gaming experience is described as ideal.