The Preferred - You Might Also Have:
Core data science foundations
• 5+ years building end-to-end predictive models in production: from raw data through feature engineering, model training, evaluation, and deployment
• Applied statistics: hypothesis testing, Bayesian methods, time-series modeling, uncertainty quantification, and understanding of common ML evaluation failure modes
• Proficiency in Python (pandas, scikit-learn, PyTorch or equivalent); advanced SQL; familiarity with cloud data platforms (AWS, GCP, or Azure)
AI agent and RAG data experience
• Direct experience building datasets and evaluation pipelines for conversational AI, chatbot, or agent systems
• Understanding of how predictive model outputs (scores, probabilitie