Controlled and Compliant AI Applications - Part 2
Explore practical methodologies for developing consistent, structured output from compliant AI systems in this 25-minute conference talk by Daniel Whitenack at the LLMs in Production Conference. Learn how to build robust systems that address concerns of corporate lawyers, finance departments, and security professionals regarding hallucinations, costs, compliance issues, data protection, and vulnerabilities. Discover techniques for integrating open access models and LLM wrappers to create AI applications that delight customers while navigating increasing restrictions on GPT models. Gain insights from Whitenack's extensive experience in developing and deploying machine learning models at scale, and learn about his work on Prediction Guard, a tool designed to simplify the developer experience when working with state-of-the-art models.