Research Engineer – World Modelling

MBZUAI
Anywhere
Workplace: RemoteFull timeFunction: Research & Scientific (R&D)Education: mastersSkills: ["Problem-solving","Troubleshooting","Communication","Collaboration"]

Design and maintain scalable video data pipelines for world model training, including preprocessing and large-scale annotation workflows. Build and optimize distributed training systems for multimodal foundation models, improving inference and serving efficiency for real-time interaction. Monitor performance and debug issues across the training/inference stack, while integrating research prototypes into production-ready, reproducible infrastructure. Contribute to quantitative evaluation tooling using rigorous benchmarks.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
MBZUAI
MBZUAI
5 months ago

Research Engineer – World Modelling

✓ Verified Job

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

Source: Company careers pageValidated by: Fursa AI
Last checked: 10 hours agoStatus: Live
Reposted: similar role first listed 1 month ago

Job Summary

Design and maintain scalable video data pipelines for world model training, including preprocessing and large-scale annotation workflows. Build and optimize distributed training systems for multimodal foundation models, improving inference and serving efficiency for real-time interaction. Monitor performance and debug issues across the training/inference stack, while integrating research prototypes into production-ready, reproducible infrastructure. Contribute to quantitative evaluation tooling using rigorous benchmarks.
Location: Anywhere
Workplace: Remote
Employment Type: Full time
Job Function: Research & Scientific (R&D)
Seniority: Mid level

Key Responsibilities

  • •Design, implement, and maintain scalable video data pipelines to support large-scale world model training.
  • •Develop preprocessing, transformation, synthesis, and large-scale data annotation workflows to ensure high-quality labels.
  • •Build and optimize distributed training systems for multimodal foundation models and improve inference efficiency for real-time interaction.
  • •Monitor system health and performance, and debug/optimize issues across the training and inference stack at scale.
  • •Integrate research prototypes into production-ready, reproducible infrastructure and support evaluation tools for quantitative model performance benchmarks.

Key Requirements

  • •MSc or PhD in Machine Learning or Computer Science, or equivalent industry experience.
  • •Experience designing robust web-scale data pipelines (data collection, cleaning, transformation) for multimodal datasets such as video/audio/text.
  • •Hands-on experience with web scraping/crawling frameworks like scrapy, selenium, playwright, or BeautifulSoup.
  • •Experience training large-scale models (LLMs or diffusion/video generative models) on large clusters and optimizing diffusion inference.
  • •Strong systems engineering skills in deep learning frameworks such as PyTorch, with the ability to debug complex training/inference failures.
Education:Master's in Machine Learning or Computer Science
Skills:Problem-solvingTroubleshootingCommunicationCollaboration
Tech Stack:PyTorchScrapySeleniumPlaywrightBeautifulSoup

Company Brief

MBZUAI
Mohamed bin Zayed University of Artificial Intelligence is a graduate-level research university in Abu Dhabi focused on AI education, research, and industry collaboration. It offers advanced programs in machine learning, computer vision, natural language processing, and related fields.
Industry: Higher Education
Company Size: Large (251 to 1,000 employees)
Growth: Government & Public Sector
Funding: Government Funded
Headquarters: Abu Dhabi, United Arab Emirates
Founded: 2019
WebsiteLinkedIn