Data Scientist

Position Overview

Use statistics and machine learning to improve training-data quality, evaluation insight, and product decisions for our video generation models. You will turn generation failures, data bias, and user behavior into actions for Research and Product — rather than owning the underlying data-pipeline engineering.

Key Responsibilities

• Define and track data-quality metrics for video / text / multimodal training corpora; identify bias, duplication, and long-tail gaps

• Design and analyze experiments and A/B tests for model versions, prompting strategies, and product features

• Analyze generation failure modes (artifacts, motion, consistency, semantic drift) and convert findings into data-collection or model-iteration recommendations

• Build analyses and dashboards that Research, Evaluation, and Product teams can act on directly

• Partner with model-evaluation engineers to calibrate automatic metrics against human ratings

• Mine API / product usage patterns to prioritize data collection and product iteration

• Communicate statistical conclusions, confidence, and business implications clearly to cross-functional partners

Qualifications & Requirements

Required Qualifications

• Master's degree in Statistics, Computer Science, Data Science, or a related field, or a Bachelor's degree plus equivalent research / industry experience

• 3+ years of data science or quantitative analysis experience

• Strong Python and SQL; comfortable with pandas, statistical modeling, and familiarity with PyTorch or a similar ML framework

• Experience in experiment design and A/B testing; understand significance, power, and common causal-inference pitfalls

• Ability to turn ambiguous product or research questions into testable hypotheses and metrics

• Clear written and verbal communication; able to explain results to non-statistical partners

Preferred Qualifications

• Experience analyzing large-scale, multimodal, or video datasets

• Familiarity with AIGC / video generation products, or prior work in model evaluation or data curation

• Experience building experiment platforms, feature analyses, or data-quality monitoring

• Bilingual proficiency in English and Chinese

What We Offer

• The chance to work on transformative AI technology and build infrastructure that powers cutting-edge products

• Competitive salary and equity package

• Opportunities for professional growth and skill development

• Flexible working arrangements (remote or on-site)

• Comprehensive health and wellness benefits

To Apply:

Please send your CV, Cover Letter, and Portfolio to [email protected] with the subject line "Data Scientist - [Your Name]".