Applied Researcher 2 - Multimodal AI

Role and team

What you will accomplish

• Develop and deploy computer vision or multimodal ML models for search, recommendations, video, and live streaming.
• Own applied ML projects from problem framing and data preparation through experimentation, evaluation, implementation, and productionization.
• Design and improve training, inference, and evaluation pipelines, including workflows orchestrating multiple models and services.
• Use metric design, offline and online evaluation, error analysis, and ablation studies to support rigorous model development.
• Contribute to detection, segmentation, classification, contrastive learning, visual-language modeling, and multimodal retrieval.
• Build agentic or VLM-powered pipelines combining foundation models, retrieval, tools, and business logic.
• Partner across product, engineering, design, and research; communicate progress, trade-offs, and results.

Qualifications

• Ph.D. or M.S. in a relevant technical field focused on computer vision, multimodal learning, or artificial intelligence.
• At least 3 years building computer vision and/or multimodal systems for large-scale applications.
• Strong knowledge of computer vision or multimodal learning techniques, including detection, segmentation, classification, representation learning, and contrastive learning.
• Hands-on experience training, fine-tuning, evaluating, and deploying visual-language models or other foundation-model-based systems.
• Practical experience with agentic AI workflows and/or VLM-powered pipelines in applied or production settings.
• Experience with rigorous experimentation, evaluation methodology, and data-driven iteration.
• Proficiency in Python and modern ML/data frameworks such as PyTorch and TensorFlow; fluency in agentic coding and AI-assisted development.
• Strong written and verbal communication and cross-functional collaboration skills.

Compensation

• Expected base pay range is C$142,400 to C$190,100; offer may vary with location, skills, and experience.
• Total compensation may include a target bonus, restricted stock units, and benefits.