Purchase Required

You need to purchase this content in order to view it

MLOPs and Guardrails Day 1 Lab

Module
62 mins

Description

In this video, Zach walks through the setup of guardrails for his MLOPS project, focusing on the database agent. He explores how to manage toxic inputs and ensure that the AI remains respectful during user interactions. Zach demonstrates the middleware functionalities, including toxicity detection and on-topic checks, while emphasizing the importance of keeping the agent aligned with relevant data tables. He also highlights the need for the middleware to adapt dynamically as new tables are created in the database.