Data Science Intern - Influencer Marketing Platform

About the role

Tencent's Level Infinite, its global gaming brand, is hiring an Associate Data Scientist intern to build AI-powered capabilities for its influencer marketing platform. The role turns data and machine learning into product features, from creator discovery and content understanding to campaign optimization and performance prediction, working at the intersection of data science, product and engineering.

Key responsibilities

• Develop and deploy machine learning and AI models, including creator segmentation and recommendation, content labeling and analysis, and campaign performance prediction and optimization
• Translate business and product problems into data science solutions, design experiments to validate hypotheses and iterate on models
• Work with data engineers to build scalable data pipelines, clean and structure creator and campaign data, and design features that improve model performance
• Collaborate with product managers and engineers to integrate models into the platform, including API design, model serving and inference workflows
• Monitor model performance with accuracy, precision, recall and business KPIs, maintain documentation and ensure reproducibility

Required profile

• Strong foundation in machine learning and statistics
• Ability to turn ambiguous business problems into data science solutions, with analytical, hypothesis-driven thinking
• Understanding of how models create value in a product context
• Ability to work cross-functionally with product, engineering and business teams, and to explain complex models simply
• Nice to have: experience in influencer marketing, advertising or creator platforms, familiarity with social media data (YouTube, Twitch, TikTok), and basic knowledge of data engineering or MLOps workflows

Required skills

• Python and common ML libraries such as PyTorch and scikit-learn
• Data processing with pandas and SQL
• Familiarity with LLMs, NLP or recommendation systems is a strong plus
• Understanding of model deployment (APIs, batch and real-time inference) is preferred