PhD Research Scientist Intern - Edge AI

Canva
London
Workplace: HybridInternshipFunction: Data Science & Machine LearningEducation: phdSkills: ["Documentation","Research execution"]

Work on efficient on-device deployment for video-capable vision-language models, targeting high-traffic consumer phones. You’ll explore and optimize candidate VLMs through techniques like quantization, pruning, distillation, compilation, and fine-tuning for intelligent caption placement. Build and benchmark an on-device prototype using real hardware measurements, compare against alternative optimization paths, and package results into a patent filing and paper-ready documentation.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Canva
Canva
5 hours ago

PhD Research Scientist Intern - Edge AI

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 5 hours agoStatus: Live

Job Summary

Work on efficient on-device deployment for video-capable vision-language models, targeting high-traffic consumer phones. You’ll explore and optimize candidate VLMs through techniques like quantization, pruning, distillation, compilation, and fine-tuning for intelligent caption placement. Build and benchmark an on-device prototype using real hardware measurements, compare against alternative optimization paths, and package results into a patent filing and paper-ready documentation.
Location: London
Workplace: Hybrid
Employment Type: Internship
Job Function: Data Science & Machine Learning
Seniority: Intern level

Key Responsibilities

  • •Survey video-capable VLMs and select the best starting point.
  • •Apply model optimization techniques to specialize vision-language models for on-device deployment, including quantization, pruning, distillation, compilation, and task-specific fine-tuning.
  • •Deploy and iterate the optimization-deployment loop on real high-traffic mobile hardware using the on-device inference library.
  • •Run comparative evaluation against alternative optimization paths and perform human evaluation against a server-side caption quality bar.
  • •Document findings for technical follow-through and compile outputs into a patent filing and paper publication.

Key Requirements

  • •Strong Python and hands-on PyTorch experience, including training and fine-tuning vision-language models.
  • •A solid understanding of modern vision-language and multimodal architectures, with the ability to pick up a recent paper and reproduce it.
  • •Experience with optimization methods like quantisation, pruning, or distillation, including their impact on accuracy.
  • •Experience deploying models on-device or at the edge with runtimes like Core ML, LiteRT/TFLite, ONNX Runtime, or ExecuTorch, within memory and latency budgets.
  • •Current enrolment in a PhD in ML, CS, or a related field, with first-author papers at venues such as CVPR, NeurIPS, ICCV/ECCV, ICLR, or ICML.
Experience:Edge AIMultimodalComputer visionOn-device MLEfficient ML
Education:PhD / Doctorate in ML, CS, or related field
Skills:DocumentationResearch execution
Languages:English
Tech Stack:PythonPyTorchQuantizationPruningDistillationCore MLLiteRT/TFLiteONNX RuntimeExecuTorchGemmaQwen-VLSmolVLMMiniCPM-VVision-language modelsMultimodal architecturesHardware-specific compilation

Company Brief

Canva
Canva is an online visual communications platform that provides drag-and-drop design tools, templates, and media assets for creating presentations, social media graphics, videos, documents and more, serving individuals and large enterprises globally.
Industry: Enterprise Software
Company Size: Enterprise (1,001+ employees)
Revenue: USD 1B+
Growth: Scaleup
Valuation: Decacorn (USD 10B+)
Funding: Series E+
Headquarters: Surry Hills, New South Wales, Australia
Founded: 2012
Glassdoor
Glassdoor: 3.9
WebsiteLinkedInGlassdoor