MLOps Summit MLOps Summit schedule

Where are we With AutoML?

AutoML aims to help everyone achieve state-of-the-art AI for their specific problems. Techniques such as Neural Architecture Search (NAS) push the boundary of finding efficient and high quality models. How is AutoML being used today and where is the field headed in the future? This talk gives an overview of AutoML techniques, exploring how they work and the challenges of applying them across different AI tasks and settings.

Adam Kraft is a machine learning engineer on the Google Brain Team, working on AutoML for a wide variety of AI tasks. Before Google, Adam spent eight years in computer vision and machine learning, working with satellite imagery at Orbital Insight and helping customers shop with their camera phones at Amazon.

Making Code and Humans GPU-Capable at Mailchimp

What happens when you have a bunch of data scientists, a bunch of new and old projects, a big grab-bag of runtime environments, and you need to get all those humans and all that code access to GPUs? Come see how the ML Eng team at Mailchimp wrestled first with connecting abstract containerized processes to very-not-abstract hardware, then scaled that process across tons of humans and projects. We’ll talk through the technical how-to with Docker, Nvidia, and Kubernetes, but all good ML Engineers know that wrangling the tech is only half the battle and the human factors can be the trickiest part.

3 Key Takeaways • An overview of the call stack from container, orchestration framework, OS, and all the way down to real GPU hardware • How ML Eng at Mailchimp provides GPU-compatible dev environments for many different projects and data scientists • An experienced take on how to balance data scientist’s human needs against heavy system optimization (spoiler alert: favor the humans)

Emily May Curtin is a Senior Machine Learning Platform Engineer at Mailchimp, which is definitely what she thought she’d be doing back when she went to film school. She combines her wealth of experience in DevOps, data engineering, distributed systems, and “cloud stuff” to enable Data Scientists at Mailchimp to do their best work. Truthfully, she’d rather be at her easel painting hurricanes and UFOs. Emily lives (and paints) in her hometown of Atlanta, GA, the best city in the world, with her husband Ryan who’s a pretty cool guy.

Models as Business Assets

Korri Jones is a Sr Lead Machine Learning Engineer and Innovation Coach at Chick-fil-A, Inc. in Atlanta, Georgia where he is focused on MLOps. Prior to his work at Chick-fil-A, he worked as a Business Analyst and Product Trainer for NavMD, Inc., was an adjunct professor at Roane State Community College, and instructor for the Project GRAD summer program at Pellissippi State Community College and the University of Tennessee, Knoxville. His accolades are just as diverse, and he was in the inaugural 40 under 40 for the University of Tennessee in 2021, Volunteer of the year with the Urban League of Greater Atlanta with over 1000 hours in a single calendar year and has received the “Looking to the Future” award within his department at Chick-fil-A among many others, including best speaker awards in business case competitions.

MLOps/Trusted AI Summit: Closing General Session: Complexity vs Simplicity in ML and AI Projects

Women in AI Reception: Pivoting into AI

Supreet is an AVP at Morgan Stanley. Prior to Morgan Stanley, she was a management consultant at ZS Associates where she automated different workflows and built data driven solutions for fortune 500 clients. She is extremely passionate about technology and AI and hence started her own community called DataBuzz where she engages the audience by sharing the latest AI and Tech trends and also mentors people who want to pivot in this field.

Selecting ML Algorithms and Validating

ML Practitioners often have a dilemma in identifying the right ML Model for the problem space. In this talk, I will be going over the common questions that will help in narrowing down the right next step. Developed model will have to meet certain validation metrics. The next common question is how the validation metrics proposed by scientists will have to be explained to business leaders and help them decide if the model is eligible to deployed. The next step is to find mechanisms to develop and study the online validation metrics. Often online metrics of an ML model launched will require studying the results in Treatment-Control fashion. In this talk, I will describe common development practices that helps in A/B testing of experiments.

Lakshmi is an Applied Scientist with Amazon.She has been working with Amazon Machine Learning teams for the last 4.5 years. She had the chance to be part of Alexa's NLP team, Behavior Analytics (a causal Inference division in Amazon) and Amazon Music teams (improving the voice experience in Alexa).

Shilpi Agarwal is a Data Philanthropist, Adjunct Faculty at Stanford and MIT $100K Launch Mentor.

Armed with the technical skills from her Bachelor of Engineering in Computer Science, design thinking skills from her Masters in Design, combined with 20+ years of Business and Marketing know-how by working as a Marketing Consultant for some really big and some small brands, Shilpi started DataEthics4All, troubled with the unethical use of data around her on social media, in business and in political campaigns.

DataEthics4All is a Community bringing the STEAM in AIᵀᴹ Movement for Youth and celebrating Ethics 1stᵀᴹ Champions of today and tomorrow pledging to help 5 Million economically disadvantaged students in the next 5 years by breaking barriers of entry in tech and creating awareness on the ethical use of data in data science and artificial intelligence in enterprise, working towards a better Data and AI World.

Ban is a Senior AI Research Scientist at Meta. She is working on democratizing Reinforcement Learning and enabling its use in the real world, spanning several application areas from compiler optimization to embodied AI. Ban and her team are developing ReAgent; an end-to-end platform for applied RL, checkout open source version at https://reagent.ai/

Selecting ML Algorithms and Validating

ML Practitioners often have a dilemma in identifying the right ML Model for the problem space. In this talk, I will be going over the common questions that will help in narrowing down the right next step. Developed model will have to meet certain validation metrics. The next common question is how the validation metrics proposed by scientists will have to be explained to business leaders and help them decide if the model is eligible to deployed. The next step is to find mechanisms to develop and study the online validation metrics. Often online metrics of an ML model launched will require studying the results in Treatment-Control fashion. In this talk, I will describe common development practices that helps in A/B testing of experiments.

Lakshmi is an Applied Scientist with Amazon.She has been working with Amazon Machine Learning teams for the last 4.5 years. She had the chance to be part of Alexa's NLP team, Behavior Analytics (a causal Inference division in Amazon) and Amazon Music teams (improving the voice experience in Alexa).

Jerry Xu is the Architect for Business Integrity's Machine Learning Infrastructure at Facebook. Jerry has over 20 years of experience building high-performance and large-scale systems with multiple engineering and leadership positions at Lyft, Box, Twitter, Zynga, and Microsoft. Before joining Facebook, he was the co-founder and CEO at Datatron Technologies, a pioneer in MLOps automation. Jerry holds several patterns on machine learning and data storage.

Taming Signals, Features, and Training Datasets: How data management shaped Pinterest ML

This talk examines how Pinterest evolved its handling of three types of data -- raw signals, ML features and ML training datasets -- and the effects on ML practitioners at Pinterest. Data management is the core complexity of production ML engineering, especially in Web-scale applications with billions of entities and training examples. Pinterest “signals,” or raw data about Pins, boards, and other entities, started as monolithic datasets that grew unmaintainable. We split them into individually owned datasets on a standardized “Signal Platform,” improving governance around lineage, ownership, and monitoring. We standardized ML features from highly custom formats to a flat “Unified Feature Representation,” enabling a shared feature store and model inference. Finally, we are transitioning ML training datasets from ad-hoc row-oriented datasets to standardized columnar table groups, enabling improved storage efficiency and shared training pipelines in the future.

David is the Head of ML Platform at Pinterest, which comprises ML Data, ML Training, and ML Serving teams. These teams provide infrastructure for 200 engineers and data scientists for applications spanning ads, recommendations, search, and trust/safety, handling billions of events per day. Previously at Pinterest, David also started the Related Pins recommendations and visual search teams and built one of the first ML-based recommender systems at Pinterest. He holds a bachelor's and master's degree in computer science from Stanford.