Applied AI Intern: Multimodal Benchmarking for Gaming

Razer (Asia-Pacific) Pte. Ltd; Singapore; On-site; SGD 20,000 - 33,000; Full time

Job summary

Razer (Asia-Pacific) Pte. Ltd. is seeking an Applied AI Intern to support QAC Vision and Agentic AI features. You will evaluate multimodal models and agentic workflows for gaming use cases, collaborating with AI, software engineering, and product teams to validate solutions for product adoption.

The role involves building evaluation workflows, conducting error analysis, and translating results into actionable recommendations for feature development in a fast-paced, international environment.

Qualifications

• Understanding of core AI/ML concepts, including evaluation and benchmarking
• Exposure to computer vision, multimodal AI, or generative AI
• Strong programming in Python; SQL experience
• Experience with backend development using NodeJS or FastAPI is a plus
• Knowledge of cloud technologies and containerisation tools (Docker/Kubernetes) is a plus
• Strong analytical and problem-solving skills with curiosity to question assumptions

Responsibilities

• Evaluate vision, multimodal, and agentic AI approaches for QAC use cases
• Build and optimise evaluation workflows for model quality and agent performance
• Develop benchmarking utilities and prototypes for technical validation
• Document findings, limitations, and failure modes with actionable recommendations
• Conduct research on emergent AI and GenAI technologies for potential adoption

Skills

Python

SQL

NodeJS

FastAPI

Docker

Kubernetes

Analytical thinking

Education

Pursuing CS/AI degree

Tools

Docker

Kubernetes

Job description

Razer (Asia-Pacific) Pte. Ltd. is seeking an Applied AI Intern to support QAC Vision and Agentic AI features. You will evaluate multimodal models and agentic workflows for gaming use cases, collaborating with AI, software engineering, and product teams to validate solutions for product adoption.

The role involves building evaluation workflows, conducting error analysis, and translating results into actionable recommendations for feature development in a fast-paced, international environment.