Senior Research Engineer - Video

• Build the next generation of generative video technology at Canva, pushing the boundaries of AI-driven storytelling. • Develop and ship cutting-edge video generation models that help hundreds of millions of people create and tell stories visually. • Partner with Research Scientists on multimodal generative AI research, translating theory into practical applications. • Work from Canva's London campus in Hoxton Square, Shoreditch, with choice in where and how you work.

• Have deep experience developing generative AI models, including Diffusion Models, GANs, or Transformers, and speak to their real-world application. • Have successfully managed and optimised large-scale distributed training across hundreds of GPUs and understand the infrastructure trade-offs. • Bring a strong understanding of ML principles and have used frameworks like PyTorch to develop and optimise performant models. • Demonstrate solid engineering practices — clean code, rigorous testing, CI/CD workflows, and robust observability in production. • Thrive in ambiguity, show end-to-end ownership of complex initiatives, and consistently drive for pragmatic solutions. • Communicate clearly, collaborate with kindness, and value knowledge-sharing and co-creation across teams.

• Have specific experience with video generation models or temporal generative architectures. • Bring hands-on experience building sustainable ML pipelines that support continuous integration, deployment, and monitoring of generative models in production. • Have experience with multimodal learning across text, image, and video modalities. • Have worked on dataset curation, annotation pipelines, and data quality for training large-scale generative models. • Bring experience with production incident investigation, root cause analysis, and sustainable fixes for ML systems.

• Partner with Research Scientists on research in multimodal generative AI, translating theory into practical applications. • Port experimental Python code to scalable, testable, and production-grade solutions, incorporating performance benchmarks and robust testing practices. • Design, develop, and implement innovative model architectures and algorithms for multimodal design generation. • Build sustainable ML pipelines that support continuous integration, deployment, and monitoring of generative models in production environments. • Optimise and scale models for efficiency, latency, and throughput across large distributed systems. • Enhance and maintain high-quality datasets and annotations to fuel multimodal learning. • Investigate and address production incidents through root cause analysis and sustainable fixes. • Collaborate closely with cross-functional stakeholders across Canva to build aligned, technically feasible, and high-impact solutions.

• Receive equity packages to share in Canva's success. • Access an inclusive parental leave policy that supports all parents and carers. • Use an annual Vibe & Thrive allowance to support your wellbeing, social connection, home office setup, and more. • Take advantage of flexible leave options that empower you to recharge and support you personally. • Work on one of the most exciting product frontiers in AI, helping redefine how people create and tell stories visually.

Canva is a global design platform redefining how the world experiences design. The Video team sits within Engineering, building next-generation generative video technology that helps Canva's hundreds of millions of users go from idea to finished video with tools that feel genuinely magical. The team works at the cutting edge of multimodal generative AI, developing and shipping video generation models that shape one of the most exciting product frontiers in AI.