Bringing ML models into production is hard. Fortunately, community initiatives such as apply(conf) are helping.
Bringing ML models into production is hard. Teams still struggle with issues such as communication between the different stakeholders involved in the ML lifecycle and data scientists being tasked with bringing models into production despite their lacking the necessary software engineering skills.
To solve these challenges and others, data and ML practitioners need to come together to share best practices and frameworks to help each other successfully build and manage production ML applications, said Gaetan Castelein, VP Marketing at ML feature store company Tecton, in an interview with RTInsights.
Tecton is helping make that happen through different community initiatives in recent years, including this week’s second apply(conf) on May 18-19, where “everything is on the table from managing labeling pipelines to transforming features in real-time to serving at scale,” according to the event announcement.
From zero to 10,000
Interest in sharing practical knowledge has been strong in the MLOps space. MLOps Community, an active community of practitioners that is sponsored by Tecton and other companies in the space and led by Demetrios Brinkmann who will MC apply(conf), grew from zero to over 10,000 in the past two years, Castelein noted.
At Tecton’s first apply(conf) last year, there were over 7,000 registrants and even more are expected this year, he continued.
“There’s a hunger for knowledge,” Castelein said.
Topics that will be covered at this year’s apply(conf) include best practices development patterns, tooling and infrastructure of choice, managing labeling pipelines, transforming and serving features in real-time, and serving at scale.
Highlighted Day 1 talks
8:35 am: Managing the Flywheel of ML Data, Mike Del Balso, Co-Founder & CEO, Tecton
9:00 am: Lakehouse: A New Class of Platforms for Data and AI Workloads, Matei Zaharia, Co-Founder and Chief Technologist, Databricks
9:35 am: Machine Learning Platform for Online Prediction and Continual Learning, Chip Huyen, Co-Founder & CEO, Claypot AI
12:00 pm: How to Draw an Owl and Build Effective ML Stacks, Sarah Cantazaro, General Partner, Amplify Partners
12:35 pm: Panel: What Do Engineers Not Get About Working with Data Scientists?
Kevin Stumpf, Co-Founder and CTO, Tecton; Maria Cipollone, User Researcher, Spotify; Mark Freeman, Founder, On The Mark Data; Kate Kuznecova, Data Scientist , OLX Group; Alice Jacques, Head of Data Science and Machine Learning, Depop
1:10 pm: Is Production RL at a tipping point? Dr. Waleed Kadous, Head of Engineering, Anyscale
Highlighted Day 2 talks
8:35 am: Empowering Small Businesses with the Power of Tech, Data, and Machine Learning Daniele Perito, Co-Founder & Chief Data Officer, Faire
9:10 am: Weaver: CashApp’s Real Time ML Ranking System, Meenal Chhabra, Machine Learning Engineering Manager, CashApp and Austin Mackillop, Machine Learning Engineer, CashApp
9:45 am: Feature Engineering at Scale with Dagger and Feast, Ravi Suhag, VP of Engineering, Gojek
12:45 pm: Intelligent Customer Preference engine with real-time ML systems, Manoj Agarwal, Architect Fellow, Walmart Global Tech and Praveen Kumar Kanumala, Principal Software Engineer, Walmart Global Tech
Learn more and register for the free event here.
