SQL or Python? Why It's the Wrong Question in the AI Era

Explore how data infrastructure and model infrastructure are converging in this 15-minute conference talk from Ray Summit 2025, where Benjamin Wagner from Firebolt Analytics examines the evolution of GenAI-first data systems. Discover why the traditional debate between SQL and Python frameworks is becoming obsolete as AI workloads increasingly demand integrated approaches to training, retrieval, transformation, and serving. Learn about the emerging distributed compute patterns that are blurring architectural boundaries between SQL engines and Python-native data frameworks, and understand what each approach contributes to modern GenAI pipelines—from SQL's query optimization and predictable performance to Python's flexibility and seamless ML ecosystem integration. Gain insights into how teams are now using unified clusters for ETL, vector search, training-adjacent processing, and model-serving workloads, and prepare for the future of unified data and ML infrastructure where ideas from both worlds are shaping the next generation of AI-driven data systems.