Bringing Feature Stores and MLOps to the Enterprise at Tecton

22 snips

Jan 5, 2021

Kevin Stumpf, Co-founder and CTO of Tecton, discusses the evolution of feature stores and their essential role in modern machine learning ops. He shares insights from his experience with Uber's Michelangelo platform and explains how Tecton simplifies feature creation for data scientists. Topics include the architecture of Tecton, the importance of observability in data management, and the challenges of integrating machine learning workflows. Stumpf also touches on the balance between open-source and enterprise solutions in the ever-evolving data landscape.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

ANECDOTE

Michelangelo Platform Impact

Kevin Stumpf led Uber's Michelangelo platform that enabled deploying thousands of ML models in production.
It powered key applications like Uber's ETAs and Eats recommendations, showcasing ML's operational scale.

INSIGHT

Operational ML Complexity

Operational ML is complex because it tightly integrates application, model, and data (features).
Feature stores solve the hardest problem: managing raw data transformations into production-ready features.

INSIGHT

Feature Definition Clarity

Features are aggregated, derived signals crucial for ML predictions, not just raw data points.
For example, frequently ordered cuisine type forms a richer predictive feature than simple click or order events.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Summary

As more organizations are gaining experience with data management and incorporating analytics into their decision making, their next move is to adopt machine learning. In order to make those efforts sustainable, the core capability they need is for data scientists and analysts to be able to build and deploy features in a self service manner. As a result the feature store is becoming a required piece of the data platform. To fill that need Kevin Stumpf and the team at Tecton are building an enterprise feature store as a service. In this episode he explains how his experience building the Michelanagelo platform at Uber has informed the design and architecture of Tecton, how it integrates with your existing data systems, and the elements that are required for well engineered feature store.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management
When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $60 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show!
Do you want to get better at Python? Now is an excellent time to take an online course. Whether you’re just learning Python or you’re looking for deep dives on topics like APIs, memory mangement, async and await, and more, our friends at Talk Python Training have a top-notch course for you. If you’re just getting started, be sure to check out the Python for Absolute Beginners course. It’s like the first year of computer science that you never took compressed into 10 fun hours of Python coding and problem solving. Go to dataengineeringpodcast.com/talkpython today and get 10% off the course that will help you find your next level. That’s dataengineeringpodcast.com/talkpython, and don’t forget to thank them for supporting the show.
You invest so much in your data infrastructure – you simply can’t afford to settle for unreliable data. Fortunately, there’s hope: in the same way that New Relic, DataDog, and other Application Performance Management solutions ensure reliable software and keep application downtime at bay, Monte Carlo solves the costly problem of broken data pipelines. Monte Carlo’s end-to-end Data Observability Platform monitors and alerts for data issues across your data warehouses, data lakes, ETL, and business intelligence. The platform uses machine learning to infer and learn your data, proactively identify data issues, assess its impact through lineage, and notify those who need to know before it impacts the business. By empowering data teams with end-to-end data reliability, Monte Carlo helps organizations save time, increase revenue, and restore trust in their data. Visit dataengineeringpodcast.com/montecarlo today to request a demo and see how Monte Carlo delivers data observability across your data infrastructure. The first 25 will receive a free, limited edition Monte Carlo hat!
Your host is Tobias Macey and today I’m interviewing Kevin Stumpf about Tecton and the role that the feature store plays in a modern MLOps platform

Interview

Introduction
How did you get involved in the area of data management?
Can you start by describing what you are building at Tecton and your motivation for starting the business?
For anyone who isn’t familiar with the concept, what is an example of a feature?
How do you define what a feature store is?
What role does a feature store play in the overall lifecycle of a machine learning project?
How would you characterize the current landscape of feature stores?
What are the other components that are necessary for a complete ML operations platform?
At what points in the lifecycle of data does the feature store get integrated?
What types of data can feature stores manage? (e.g. text vs. image/binary vs. spatial, etc.)
How is the Tecton platform implemented?
- How has the design evolved since you first began building it?
  - How did your work on Uber’s Michelangelo inform your work on Tecton?
What is the workflow and lifecycle of developing, testing, and deploying a feature to a feature store?
What aspects of a feature do you monitor to determine whether it has drifted?
- How do you define drift in the context of a feature?
  - How does that differ from drift in an ML model?
How does Tecton handle versioning of features and associating those different versions with the models that are using them?
What are some of the most interesting, innovative, or unexpected projects that you have seen built with Tecton?
When is Tecton the wrong choice?
What do you have planned for the future of the product?

Contact Info

LinkedIn
kevinstumpf on GitHub
@kevinstumpf on Twitter

Parting Question

From your perspective, what is the biggest gap in the tooling or technology for data management today?

Closing Announcements

Thank you for listening! Don’t forget to check out our other show, Podcast.__init__ to learn about the Python language, its community, and the innovative ways it is being used.
Visit the site to subscribe to the show, sign up for the mailing list, and read the show notes.
If you’ve learned something or tried out a project from the show then tell us about it! Email hosts@dataengineeringpodcast.com) with your story.
To help other people find the show please leave a review on iTunes and tell your friends and co-workers
Join the community in the new Zulip chat workspace at dataengineeringpodcast.com/chat

Links

The intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA

Support Data Engineering Podcast