Member of Technical Staff - Multi-Modal, Vision

10 months ago

H1B Sponsor

Responsibilities

Lead model capabilities end-to-end from task specification and data curation through training, ablations, evaluation, and shipment.
Improve visual reasoning using reinforcement learning and preference optimization methods.
Advance token efficiency through encoder and connector design while preserving model quality.
Own a major workstream such as video understanding, preference data quality, or encoder architecture.
Contribute directly to shipping at least one production model.

Hands-on experience training or evaluating vision-language models with demonstrated experimental rigor.
Ability to translate research ideas into scalable implementations and iterate through hypotheses.
Proficiency in Python and at least one deep learning framework.
M.S. or Ph.D. in Computer Science, Mathematics, or a related field, or equivalent industry experience.
Experience with multimodal training or data pipelines, distributed training, multimodal post-training, dataset design and data quality, computer vision, or visual representation learning is valuable.
Prior open-source contributions through code, data, or models on GitHub or Hugging Face are valued.
Research publications at leading AI conferences such as NeurIPS, ICML, CVPR, ECCV, ICLR, or ACL are valued.

Competitive base salary with equity in a unicorn-stage company
100% employer-paid medical, dental, and vision premiums for employees and dependents
401(k) matching up to 4% of base pay
Unlimited paid time off and company-wide Refill Days

51-200 employees

We build efficient general-purpose AI at every scale.