ML Research Scientist, Pegasus

📋 Description

• • As an ML Research Scientist on the Pegasus team at Twelve Labs, you will drive cutting-edge research to advance video understanding capabilities in multimodal foundation models, directly contributing to products that enable human-like comprehension of video content.
• • You will define and drive research problems, design and run rigorous experiments across model architecture, training strategies, data curation, and evaluation, build evaluation methods and data curation pipelines, collaborate with ML engineers to transition research into production systems, and communicate findings to inform technical direction.
• • The Pegasus team is at the core of Twelve Labs’ video understanding capabilities, responsible for developing the Pegasus Video Analysis product with a focus on real-world impact, high instruction following, and hierarchically structured outputs, working in a goal-oriented, cross-functional environment with access to advanced hardware like NVIDIA B300s.
• • You will have the opportunity to push the boundaries of multimodal AI, gain experience in large-scale distributed training and data-centric model development, see your research translate into real product impact, and grow as a researcher in a globally recognized AI company backed by top-tier investors and partners like NVIDIA and AWS.