Feature Store Summit
The Feature Store Summit, organized by Hopsworks, is the first conference dedicated to cutting-edge technologies that facilitate bringing machine learning models into production.
DAY 1
Jukebox : Spotify's Feature Infrastructure
The team at Spotify talk about the challenges when building a central feature infrastructure at a highly autonomous organization.
Hopsworks Feature Store: Fast, Fresh Data for AI at Scale
The Hopsworks team presents how to leverage the Feature Store to make real-time predictions, with low-latency and at scale.
Building Feature Store for Multi-tenant and Multi-App in Salesforce
The team at Salesforce discusses how the company leverages from the Feature Store to build a collaborative environment across teams.
Feature Store: The Heart of Your Operational ML Pipeline
Yaron Haviv from Iguazio talks how the feature store supports real-time and batch use cases across training and serving environments.
Twitter's management of ML features in dynamic environments
David Liu talks about how Twitter solved challenges of collaboration and shareability with a centralized feature store.
Panel Discussion - Solving the Hardest Problems at Scale
Moderated by Mike Klaczynski from Snowflake.
Spotify, Salesforce, Twitter, Hopsworks, Iguazio and Disney discuss the significant role of feature stores as a company scales.
Build models faster, and serve predictions at scale using Amazon SageMaker Feature Store
Mark Roy from Amazon talks how SageMaker can help to accelerate the ML lifecycle, providing low-latency and high throughput inference.
Creating and operating ML models from event-based data using feature stores and feature engines
The teams from Kaskada and Redis focused on how iterate on amazing ML models with event-based data.
Databricks Feature Store Co-designed with a Data and MLOps Platform
The team at Databricks talk about the motivations and use cases of feature stores across different industries.
Taming the beast: Building Scalable Features in the Wild at Prescient
The Rasgo and Prescient team will talk about the best setup, and what’s involved in getting features/models to be production-ready.
Feature Store: the heart of the MLOps framework
Richa Sachdev from Vanguard discusses the role of the feature store for MLOps for a successful analytical journey.
Panel Discussion - Who Benefits from the Feature Store?
Moderated by Nicholas Pinckernell from Comcast.
AWS, Databricks, Rasgo, Kaskada, Vanguard and Shelf Engine dive into the specific benefits for data science and MLOps.
DAY 2
Being ‘Data Centric’ is the Future of Machine Learning
Atindriyo Sanyal from Galileo talks about data centric aspects of machine learning.
Feature Stores and Evaluation Stores: Better Together
Josh Tobin from Gantry talks about the evaluation store and why to combine it with the feature store for more robust ML systems.
The SAME Project: A Cloud Native Approach to Reproducible Machine Learning
David Aronchick from Microsoft presents the Self-Assembling Machine Learning Environment, a new Kubernetes and Kubeflow project.
The Coming Wave of Self-Supervised Embedding Ecosystems
Laurel Orr from Stanford University discusses the challenges and opportunities with supporting embedding pipelines in feature store.
Why Relational Learning Matters - Automated Feature Engineering on Relational Data and Time Series
Patrick Urbanke from getML talks about how relational learning can be used to automate feature engineering, reducing time and costs.
Panel Discussion - The Future of Feature Stores
Moderated by Chip Huyen from Stanford.
Microsoft, Stanford, getML, Galileo, and Gantry. discuss some of their predictions on what will become increasingly important in the coming months and years.
A Software Development Ecosystem that Makes Developers Happy
Robert Lock from Bosch discusses why companies shouldn't build their own solution when a software already exists as PaaS.
Feature Store at Varo: Why, How and Lessons Learned
The team at Varo walks through the evolution of the feature store, from the ontological challenges to key functionalities.
A Query-Based Feature Store at OLX
Augusto Acioli from OLX presents a feature store where Data Scientists can create their online feature using queries.
Palette at Scale
The team at Uber talk about the advances of the Palette Feature Store that enables Feature Management at scale.
Feature Store at Via: Implementation, Difficulties and ROI
Cezar Steinz presents the ROI of the feature store according to the platform and models implementation at Via.
Better Gaming Experiences with Machine Learning and the Hopsworks Feature Store
Renan from Wildlife Studios discusses how the Hopsworks Feature Store is helping them scale a centralized ML platform.
Panel Discussion - Feature Store: Build or Buy?
Moderated by Jim Dowling from Hopsworks.
Uber, Bosch, Varo, OLX, Via and Wildlife Studios discuss about what to consider when buying or building a feature store for ML.