FICO, a 60-year-old company, pushes boundaries in real-time machine learning model development and management, requiring low latency and high accuracy to process thousands of transactions per second, while balancing costs and constraints. The company's Chief Analytics Officer, Scott Zoldi, shares insights on the tradeoffs between latency, analytics accuracy, and other factors when architecting customer solutions, and emphasizes the importance of real-time data processing and explainable AI in decision-making. FICO develops flexible data structures and integrates models into production environments through a collaborative effort between data scientists, software developers, and cloud engineers, while prioritizing R&D teams that explore fully explainable ML models and responsible AI practices.