Serverless & Event-Driven Patterns for GenAI

Explore serverless and event-driven architectural patterns for building Generative AI solutions in this conference talk from GOTO EDA Day 2023. Discover how to efficiently host, train, and consume GenAI using AWS services. Learn about various use cases, the benefits of serverless architecture for GenAI, and key serverless services. Dive into specific patterns for implementing GenAI solutions, including context-based tuning and document summarization at scale. Gain insights on hosting foundation models and leveraging SageMaker for large-scale processing. By the end, acquire practical knowledge on implementing event-driven architectures to create scalable and cost-effective GenAI applications.

Intro
Agenda
Introduction to generative Al
Use cases
Why use serverless with generative Al?
Serverless services
Serverless patterns for generative Al
How to build context based on internal knowledge?
Hosting foundation models
Documents summarization on SageMaker at scale
Summary
Outro