Solving the Last Mile Problem of Foundation Models with Data-Centric AI

Explore the challenges and solutions for achieving production-level accuracy with large language models in this conference talk by Alex Ratner. Learn about the "last mile" problem in deploying AI applications, particularly in enterprise settings. Discover how data-centric AI approaches can address issues such as hallucinations, data biases, and misclassification of domain-specific edge cases. Gain insights into the concept of foundation models as powerful starting points for AI development, and understand why additional steps are necessary to build robust, production-ready AI systems. Benefit from Ratner's expertise as the co-founder and CEO of Snorkel AI, and his background in developing data-centric AI techniques at Stanford University.

Solving the Last Mile Problem of Foundation Models with Data-Centric AI //Alex Ratner // LLM in Prod