In this episode of Two-Shot, host Julia welcomes Shreya Shankar, a software engineer and AI researcher pursuing her PhD at UC Berkeley's EECS department. Shreya, known for her work on machine learning infrastructure and LLM evaluation, shares insights from her research on human-centered, AI-powered data processing. From testing frameworks to evaluation metrics, they explore the practical challenges of assessing large language models and discuss innovative solutions for developers working with these systems. This episode offers valuable perspective on the evolving landscape of AI evaluation and deployment, making it essential listening for ML practitioners and technology enthusiasts.