Are you a Gen AI/ML practitioner, data scientist or a business leader struggling to decide between customizing foundation models or augmenting them with external knowledge? Join us for an immersive hands-on session that tackles the critical question in Gen AI implementation: when to implement Retrieval-Augmented Generation (RAG) versus when to fine-tune your models.

In this workshop, you'll get practical experience with both approaches using the AWS comprehensive suite of Gen AI/ML services including Amazon Bedrock, Amazon SageMaker, and model evaluation tools. By the end of the session, you'll have a clear decision framework to guide your organization's model customization strategy.

Who is this for? This workshop is ideal for:

• AI/ML practitioners and engineers implementing Gen AI solutions.
• Technical and Business leaders evaluating model customization approaches.
• Solution architects designing knowledge-intensive applications.
• Developers working with context-rich Gen AI use cases.
• Business stakeholders looking to understand tradeoffs in AI customization.

Key highlights:

• Demo of RAG pipelines using Amazon Bedrock.
• Step-by-step fine-tuning of foundation models with Amazon SageMaker and Amazon Bedrock.
• Practical evaluation frameworks to measure performance of both approaches.
• Cost-benefit analysis of RAG vs fine-tuning strategies.
• Real-world case studies illustrating optimal use cases for each approach.
• Comprehensive decision framework development for your specific business needs.
• Best practices for knowledge integration in large language models (LLMs).

This session provides a balanced perspective on both customization strategies, helping you make informed decisions based on your specific use cases, data availability, and performance requirements. You'll leave with practical implementation knowledge and a strategic framework for choosing between RAG and fine-tuning.

Prerequisites:

• Basic to intermediate understanding of LLMs.
• Familiarity with Python programming.
• Basic knowledge of vector databases and embeddings.
• Understanding of prompt engineering concepts.

By registering, you agree to the AWS Event Terms & Conditions and AWS Code of Conduct.