Jonathan Buck

An agent’s performance depends on far more than its model or system prompt. Tools, instructions, context management, memory, orchestration, and runtime policies form an agent harness that can determine whether the same model succeeds or fails.

In this chalk talk, we explain how to use customer pre- and post-Lambda functions to extend the functionality of Amazon SageMaker Ground Truth. We show you how to make a custom annotator UI and use AWS Lambda to provide custom pre- and post-processing of annotations.

Amazon SageMaker Ground Truth is a fully managed service for data labeling. In this chalk talk, we explore the tools and techniques available in Ground Truth to help generate high-quality labels, from audit and verification workflows to annotation consolidation and active learning.