Own the quality and reliability of AI features Partner with Product Managers, Engineers, and Designers Frame problems, design experiments, and make evidence-based decisions Deliver evaluation and analysis code Collaborate with engineers to productionize changes Develop a deep understanding of Ada’s product and empathize with users Run rapid, reliable experiments with sound statistical methods Lead design choices across model selection, retrieval and grounding, prompting, and tool use Reduce critical AI failure modes (e.g., hallucinations, prompt injection) Partner with PM, Engineering, and Design to prioritize customer use cases Coach and mentor peers in AI feature design, experimentation, and measurement