Video-Generation Research Scientist & ML Systems

Cantina Labs; Singapore; On-site; SGD 70,000 - 100,000; Full time

Benefits offered by this job

Competitive salary and generous company equity

Personal time off and paid holidays

Health insurance

Global travel insurance

Monthly spending stipend: $500 (~S$635)

All equipment needed for home office

Job summary

Cantina Labs in Singapore is seeking a motivated Research Scientist to lead foundational research on video generation models. This role involves ownership of the full research cycle and requires collaboration with various teams to implement model enhancements. The ideal candidate has strong experience in large-scale data systems, containerization, and generative models, along with a solid research background. Competitive salary and generous equity are part of the offered benefits.

Qualifications

• Strong experience building large-scale data systems for machine learning.
• Proven skills in distributed data processing and orchestration tools.
• Experience with Docker and Kubernetes for containerization.
• Familiarity with cloud data services and video processing tools.
• Deep knowledge of generative models and their training methods.

Responsibilities

• Build scalable systems for video data processing for model training.
• Design data pipelines for dataset generation and refreshes.
• Orchestrate workflows and monitor large data processing jobs.
• Implement containerized infrastructure using Kubernetes.
• Optimize cloud data storage for cost and efficiency.
• Develop distillation methods for video generation models.
• Build reward models to align video quality with human judgments.

Skills

Large-scale data systems or pipelines for machine learning workflows

Distributed data processing frameworks (e.g., PySpark, Ray)

Containerization and orchestration (Docker, Kubernetes)

Cloud-based data storage (AWS, GCS, Azure)

Video and media processing tools (FFmpeg, PyAV, DALI, OpenCV)

Multimodal data (video, image, text, audio)

Post-training methods for generative models

Reward modeling and preference-based fine-tuning

Proficiency in Python and machine learning frameworks (PyTorch, JAX)

Independent research and project management

Job description

Cantina Labs in Singapore is seeking a motivated Research Scientist to lead foundational research on video generation models. This role involves ownership of the full research cycle and requires collaboration with various teams to implement model enhancements. The ideal candidate has strong experience in large-scale data systems, containerization, and generative models, along with a solid research background. Competitive salary and generous equity are part of the offered benefits.