The Future of DOCUMENT AI and Transitioning to AI_EXTRACT

Are you looking to modernize how you extract data from PDFs, images, and forms and move off the legacy Document AI UI and MODEL!PREDICT API before they are decommissioned on February 28, 2026? This hands‑on webinar is designed for teams that want to keep existing document workloads running while building new, scalable pipelines on AI_EXTRACT.

We will show, step by step, how to migrate existing Document AI models into the Model Registry, then update your pipelines to call AI_EXTRACT instead of MODEL!PREDICT, referencing your current models without retraining.

After preserving legacy Document AI models, we will walk through how Snowflake is standardizing document extraction on AI_EXTRACT, an AISQL function powered by the arctic‑extract model. You’ll see how AI_EXTRACT improves accuracy and performance while moving to token‑based billing rather than GPU/compute‑time pricing.

This webinar is for data and analytics engineers, developers, and architects handling document-heavy workflows. It covers migrating from legacy Document AI to AI_EXTRACT and introduces a reusable playground for document processing in Snowflake.