Obtaining New Insights on Model Behavior with Fiddler - MLOps Meetup #110

Explore new insights on model behavior with Fiddler in this 54-minute MLOps Community Meetup talk featuring Danny Brock, Solutions Engineer at Fiddler AI. Dive into the world of monitoring, explainability, and model analysis as Danny demonstrates how MLOps Engineers and Data Scientists can instrument ML models for observability using Fiddler's Model Performance Management platform. Learn to register models, publish events, and leverage Fiddler's cloud environment on AWS to gain valuable insights into data drift, integrity, and underperforming cohorts. Discover how Fiddler addresses critical ML challenges, drives positive business outcomes, and provides a comprehensive solution for the ML lifecycle. Follow along with a detailed demo and quick start guide for implementing simple monitoring, creating projects, uploading baseline datasets, and publishing production events. Gain practical knowledge on building trust in AI and integrating Fiddler into your ML workflows.

[] Introduction to Danny Brock
[] We're streaming live on Gradual and LinkedIn!
[] Fiddler AI Overview
[] AI is a critical part of modern business
[] AI touches your entire organization
[] The Fiddler mission
[] How can Fiddler make a difference?
[] Model Performance Management
[] The Comprehensive Solution for the ML Lifecycle
[] The Fiddler Value
[] Frequently heard challenges
[] Fiddler drives three positive business outcomes
[] Under the hood
[] A deep and versatile tech stack
[] Seamless UX
[] Getting started with Fiddler
[] Demo
[] Build trust into AI
[] Simple monitoring quick start
[] Fiddler simple monitoring quick start guide
[] Recap of simple monitoring
[] Step 1: Create a new project in the environment
[] Step 2: Upload the baseline data set into the project
[] Step 3: Add information about the model
[] Backfill ground truth and monitor in a synchronous way
[] Step 4: Publish production events
[] Call publish event batch
[] Not found error
[] Run without storing data at Fiddler
[] Integration support of Fiddler