Назад
18 дней назад

Principal Research Scientist - Evaluations (AI)

310 000 - 420 000$
Формат работы
hybrid
Тип работы
fulltime
Грейд
senior
Английский
b2
Страна
UK/US/Australia +2 еще
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Principal Research Scientist - Evaluations (AI): Defining and building Canva-wide evaluation systems for generative design, image, video, audio, and agentic models with an accent on human preference, multimodal quality measurement, reward signals, and production monitoring. Focus on validating metrics against user and product outcomes, diagnosing model regressions and judge bias, and setting evaluation direction across globally distributed research teams.

Location: San Francisco; hybrid work

Salary: $310,000–$420,000 annually

Company

Canva develops design products powered by generative models and serves millions of users worldwide.

What you will do

  • Define Canva Research's long-term evaluation strategy across design, image, video, audio, and agentic workflows.
  • Build shared evaluation standards covering human preference, learned quality models, MLLM judges, reward signals, and multimodal assessment.
  • Establish evaluation gates for launch decisions and diagnose anomalous results, model regressions, and infrastructure artefacts during production training.
  • Connect evaluation metrics with user experience, product outcomes, downstream business impact, and editability.
  • Partner with Design Generation, Foundation Models, Agents, design, creator, and product teams so evaluation informs training and inference.
  • Mentor senior research scientists and engineers and represent Canva's evaluation practice to leadership and the research community.

Requirements

  • Track record of defining evaluation frameworks and standards from the ground up and creating measurement systems that influence team decisions.
  • Experience linking evaluation metrics to user or business outcomes and diagnosing divergence between them.
  • Experience converting subjective human judgement into reliable evaluation signals through rubric design, human-data pipelines, and model training.
  • Strong knowledge of multimodal generative models, including diffusion models, transformers, VLMs, and MLLMs.
  • Experience with reward modelling, preference learning, or alignment methods involving human feedback.
  • Experience setting technical direction across multiple teams in a globally distributed organisation.

Nice to have

  • Research background in human perception, psychophysics, aesthetics, or HCI.
  • Experience evaluating harm, bias, and safety alongside quality.
  • Publications in evaluation, alignment, or generative modelling.
  • Background or genuine interest in visual arts and graphic design.

Culture & Benefits

  • Flexible hybrid work between home and the office.
  • Equity package and health benefits plans.
  • 401(k) retirement plan with company contribution.
  • Inclusive parental leave for all parents and carers.
  • Annual Vibe & Thrive allowance for wellbeing, social connection, and office setup.
  • Flexible leave options and virtually conducted interviews.

Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →