Presentation: Automating Netflix ML Pipelines With Meson

Track: The Practice & Frontiers of AI

Location: Seacliff ABC

Duration: 1:40pm - 2:30pm

Day of week: Monday

Level: Intermediate

Persona: Architect, CTO/CIO/Leadership, Data Engineering, ML Engineer, Technical Engineering Manager

What You’ll Learn

  • What challenges you will face in orchestrating an ML based system
  • How to leverage multiple technologies to increase the effectiveness of your ML and Data Science efforts.

Abstract

In this talk we discuss the evolution of ML automation at Netflix and how that lead us to build Meson, an orchestration system used for many of the personalization/recommendation algorithms. We will talk about challenges we faced, and what we learned automating thousands of ML pipelines with Meson.

Interview

Question: 
What is your motivation for this talk?
Answer: 

We want to give a broader story on what it means to try to automate experience tests with machine learning; when you're dealing with tons of smart data scientists that want to try lots of things. How do you build the infrastructure that's lasting and provides value while you're getting this flood of new ideas and new technology that you need to support.

Question: 
Who should come to your talk?
Answer: 

People who are familiar with the space or anyone curious about how we do orchestration at Netflix.

Question: 
What can people come take away from this talk?
Answer: 

How you transition from the experimentation phase into the production phase. That's dealing with the issues of the day to day workflow of a data engineer. And that effort doesn't have necessarily an easily measurable metric like runtime or prediction accuracy. Those are the kinds of problems that we often have and in some ways they have a bigger impact on the ability to actually leverage ML at scale, not whether or not you can run a giant neural net.

Question: 
What keeps you up at night?
Answer: 

Are there fundamental abstractions in how we think about modeling a pipeline? Because words like workflow and pipeline are sort of thrown around very interchangeably and I stay awake at night thinking about like where are the real seams here?

When you're interacting with an orchestration system like this, oftentimes so many things are sort of intertwined and braided together that you can't really tease out why it works the way it does. What are sort of the key pieces of how to actually construct these things especially in such an evolving space?

Speaker: Davis Shepherd

ML Management @Netflix

Davis Shepherd spent 4 years building re-enforcement learning systems before working at Netflix where he now develops their next generation of ML pipeline management software.

Find Davis Shepherd at

Speaker: Eugen Cepoi

Senior Software Engineer @Netflix

Eugen Cepoi has been working on data processing systems for general ETL like purposes and machine learning for the past 4 years. Before that he spent 3 years building web applications and infrastructure for them. He now works in the personalization infrastructure team at Netflix, where he focuses on developing the ML pipeline management software.

Find Eugen Cepoi at

Similar Talks

VP of Machine Learning @CrowdFlower
Director of Vulnerability Research @Endgame
Technical Program Manager @Questback
Product Management and Marketing @Datacoral
Founding Member of the Atom Editor Team @GitHub

.

Tracks

  • Architectures You've Always Wondered About

    Architectural practices from the world's most well-known properties, featuring startups, massive scale, evolving architectures, and software tools used by nearly all of us.

  • Going Serverless

    Learn about the state of Serverless & how to successfully leverage it! Lessons learned in the track hit on security, scalability, IoT, and offer warnings to watch out for.

  • Microservices: Patterns and Practices

    Stories of success and failure building modern Microservices, including event sourcing, reactive, decomposition, & more.

  • DevOps: You Build It, You Run It

    Pushing DevOps beyond adoption into cultural change. Hear about designing resilience, managing alerting, CI/CD lessons, & security. Features lessons from open source, Linkedin, Netflix, Financial Times, & more. 

  • The Art of Chaos Engineering

    Failure is going to happen - Are you ready? Chaos engineering is an emerging discipline - What is the state of the art?

  • The Whole Engineer

    Success as an engineer is more than writing code. Hear inward looking thoughts on inclusion, attitude, leadership, remote working, and not becoming the brilliant jerk.

  • Evolving Java

    Java continues to evolve & change. Track covers Spring 5, async, Kotlin, serverless, the 6-month cadence plans, & AI/ML use cases.

  • Security: Attacking and Defending

    Offense and defensive security evolution that application developers should know about including SGX Enclaves, effects of AI, software exploitation techniques, & crowd defense

  • The Practice & Frontiers of AI

    Learn about machine learning in practice and on the horizon. Learn about ML at Quora, Uber's Michelangelo, ML workflow with Netflix Meson and topics on Bots, Conversational interfaces, automation, and deployment practices in the space.

  • 21st Century Languages

    Compile to Native, Microservices, Machine learning... tailor-made languages solving modern challenges, featuring use cases around Go, Rust, C#, and Elm.

  • Modern CS in the Real World

    Applied trends in Computer Science that are likely to affect Software Engineers today. Topics include category theory, crypto, CRDT's, logic-based automated reasoning, and more.

  • Stream Processing In The Modern Age

    Compelling applications of stream processing using Flink, Beam, Spark, Strymon & recent advances in the field, including Custom Windowing, Stateful Streaming, SQL over Streams.  

Conference for Professional Software Developers