Visual Data Curation Specialist

Cantina Labs develops social AI technologies, including real-time models that enable expressive, lifelike characters. Our work connects advances in video generation and multimodal AI with new ways for people to create, communicate and interact. These capabilities underpin our flagship platform, Cantina, and the broader experiences we are building.

We’re looking for a Visual Data Curation Specialist to join our growing research team in Singapore. You will identify and evaluate image and video sources that support the development of our video generation and multimodal models.

This role combines visual expertise, research and analytical judgment. You’ll work closely with researchers and engineers to understand data requirements, translate them into sourcing priorities, and develop a systematic approach to assessing content quality and relevance.

You’ll own sourcing and curation projects from initial discovery through review and refinement, helping the team make better decisions about the visual data used in model development.

• →

  Identify and evaluate image and video sources across creator communities, specialist websites, archives and platforms such as YouTube, TikTok, Bilibili, Instagram and Vimeo.

• →

  Partner with researchers to translate model development priorities into clear sourcing briefs, selection criteria and content categories.

• →

  Assess content across aesthetics, technical quality and research relevance, considering factors such as composition, lighting, resolution, camera movement, subject motion and visual consistency.

• →

  Build and maintain a structured source catalogue with representative examples, relevant metadata, quality assessments and clear recommendations.

• →

  Identify gaps in coverage across subjects, styles, environments and types of motion, and propose targeted sourcing approaches to address them.

• →

  Review sample content with research and engineering colleagues, incorporate feedback and refine selection criteria as requirements evolve.

• →

  Document source provenance and available usage information, flagging ownership, licensing or access questions for review.

• →

  Develop repeatable curation workflows, tagging systems and quality guidelines, and identify tools that improve the efficiency and consistency of the work.

• Strong visual judgment and the ability to explain how an image or video meets creative, technical, or research requirements.

• A background in film, photography, animation, design, visual arts, media, or a related field, demonstrated through education, professional experience, or a portfolio. Prior experience in machine learning is not required.

• Strong research skills, including the ability to investigate unfamiliar topics, assess sources, and turn findings into useful recommendations.

• A systematic approach to organizing information and confidence using spreadsheets, databases, or digital asset management tools.

• Clear communication, with the ability to translate visual observations into precise feedback for technical colleagues.

• The ability to work independently, clarify open-ended requirements, prioritize tasks, and take ownership of work quality.

• Curiosity about generative AI and the ability to learn new tools and workflows quickly. Prior machine learning experience is not required.

• Hands-on experience in photography, filmmaking, video editing, animation, or other visual production.

• Experience in content research, editorial curation, digital asset management, or archive research, including knowledge of visual content communities across platforms and cultures.

• Familiarity with dataset curation, annotation, quality assurance, or model evaluation.

• Experience with generative image or video tools, basic scripting, SQL, or workflow automation.