We're hiring a Research Data Scientist to join us at an exciting early stage. You'll play a central role in helping us turn messy, real-world industry data into AI-driven insights and products. The role is a mix of data exploration, applied ML, and experimentation with LLMs. You'll be designing and running experiments, proving what works, and collaborating closely with engineers to turn research into production-ready systems. What you will do: Wrangle large, messy datasets from legacy systems and real-world processes into structured, usable formats. Design, run, and evaluate experiments with rigorous statistical methods to validate hypotheses and measure business impact. Build and prototype ML models for forecasting, optimization, and recommendation across domains like sales, inventory, and operations. Explore and fine-tune large language models (LLMs) for practical use cases in industry workflows. Collaborate with engineers to take promising research into production, ensuring models are robust, scalable, and trustworthy. Contribute to a culture of experimentation, documenting results, sharing learnings, and driving the science behind our product decisions. What we look for: You believe data is the foundation for innovation, and love analyzing and modeling it to find unexpected insights that can drive business impact. You bring a scientific mindset (Master/PhD in math/science degree), rigorous about experimentation, evidence, and statistical reasoning. 4+ years of hands-on experience in data science, ML engineering, or applied research OR a PhD in a science plus 2+ years of experience. Comfortable designing and running experiments with proper baselines, controls, and statistical evaluation. Excited to explore and apply LLMs in practical, business-focused contexts. Thrive in a dynamic, iterative environment where things move quickly and research gets put into action. No industry knowledge required, you'll learn the domain as you go, and it's a fascinating one. Technical skillset: Must: Python, familiar with common data science/ML libraries (e.g., pandas, NumPy, scikit-learn, XGBoost, PyTorch, Hugging Face), experiment design, evaluation metrics. Nice to have: LLM fine-tuning, prompt engineering, SQL/PostgreSQL.