Bruno Pistone

Agentic workloads are multi-step, tool-calling, and retrieval-augmented, which means they stress the inference layer differently than a single prompt. In this code talk, we build three patterns live: function calling, multi-step reasoning chains, and retrieval-augmented generation with Knowledge Bases.

Go from dataset to a custom foundation model in a few hours, without the need to set up any infrastructure. In this hands-on workshop, you will fine-tune LLMs using Amazon SageMaker Serverless Model Customization, requiring no infrastructure.

This workshop covers advanced fine-tuning techniques like Supervised Fine-Tuning (SFT) and Direct Preference Optimization (DPO) using Amazon Nova recipes and Amazon SageMaker AI.

Transform general-purpose AI models into powerful, specialized solutions in this hands-on workshop. Learn practical techniques from continuous pre-training to fine-tuning to model distillation that deliver higher accuracy and cost savings.

Learn how Amazon SageMaker AI transforms model development with a unified environment for all your AI workloads - from interactive model development on IDEs to maximizing task and compute resource utilization across training and inference workloads.

In this chalk-talk, attendees learn how to build and deploy domain-specific small language models (SLMs) for automotive use-cases. Explore how Amazon SageMaker can be used to fine-tune an SLM using industry terminology, and how the SLM can be scaled using Amazon Bedrock.