Tecton 0.6: Notebook-driven Development for Feature Engineering - MLOps Meetup

Explore the latest advancements in feature engineering with this 42-minute MLOps Community Meetup talk by Jason Dunne, Senior Product Marketing Manager at Tecton. Dive into Tecton 0.6's new notebook-driven development capability, enabling data teams to develop and test features quickly using Python notebooks. Learn about real-time ML challenges, Tecton's feature platform overview, and how it streamlines the process of designing, building, centralizing, serving, and managing features for production ML. Witness a live demo of notebook-driven development and understand its benefits in reducing steps for creating new features. Gain insights into production workflows, low-latency ingestion for real-time feature pipelines, and self-hosting considerations. Perfect for data engineers, ML engineers, and data scientists looking to enhance their feature engineering workflows.

[] Musical intro for Jason Dunne
[] Notebook-driven Development
[] Tecton 0.6: Agenda
[] Real-Time ML
[] Uber Eats: Real-Time ML At Scale
[] Building Real-Time systems is hard. Maintaining them is harder
[] Tecton Feature Platform Overview
[] Tecton allows teams to quickly and reliably transform and serve data for ML applications, at scale.
[] Tecton Feature Platform: Design, Build, Centralize, Serve, and Manage Features for Production ML
[] Tecton is the Feature Platform of choice for leading ML teams across industries and use cases
[] 0.6 Core Capabilities
[] Public Preview: Notebook-driven Development
[] Notebook-driven Development reduces the number of steps for developers creating new features
[] Notebook-driven Development Demo
[] Tecton's Feature Platform
[] Feature depends on another feature
[] Production flow
[] Difference from TDD
[] Low-latency ingestion that powers real-time feature pipelines
[] Sufferance from sweet test labeling
[] Being self-hosted
[] Notebook cluster
[] Wrap up