Feature stores, and second thoughts
Feature stores promised to fill a gap in the ML platform. People who built and ran them explain where they helped, and why they sometimes reconsidered.
Mid-Scale Production Feature Engineering
Pushed backDr. Venkata Pingali rejects the idea that feature engineering is a simple linear process in which data is merely fed into a model.31:38
Running a Fintech on Machine Learning
Pushed backThe team favored building its feature store internally rather than buying an existing product because it wanted tighter integration and customization, and found no sufficiently mature market solution.46:45
Feature Stores: An Essential Part of the ML Stack to Build Great Data
Pushed backKevin Stumpf rejects the idea that Tecton should be an end-to-end machine-learning platform, arguing that a focused feature platform can integrate with best-in-class tools for other parts of the workflow.53:00
A Conversation Around Feature Stores
Pushed backVenkata Pingali said the future design space would contain many different feature-store architectures rather than one standard implementation.45:42
Real-time Feature Pipelines, A Personal History
Pushed backHendrik Brackmann says the usefulness of storing dynamic model outputs as features depends on reuse, training design, and system boundaries rather than having one universal answer.45:02
Machine Learning Feature Store Panel Discussion
Pushed backWhether a company needs to buy or build a feature store depends on its use cases and maturity rather than having a single universal answer.9:29
Machine Learning Design Patterns for MLOps
ClaimA feature store is useful when multiple models need to reuse the same calculated features, but many use cases only need a library.13:04
Building ML Blocks with Kubeflow Orchestration with Feature Store
Pushed backAniruddha Choudhury says Feast is not a workflow scheduler, data warehouse, pipeline orchestrator, feature engineering tool, or model-serving product.16:10
Feature Stores at Shopify and Skyscanner
Pushed backMatt Delacour says feature engineering is outside the feature store project, while the question concerned whether feature services contain feature-engineering code.25:26
Building a Movie Recommendation System on Tecton with Snowflake
Pushed backDavid Hershey says Tecton does not currently provide built-in automatic drift detection, while external tools are commonly used to profile feature data.47:52
Feathr: LinkedIn's High-performance Feature Store
Pushed backReal-time features should not be treated as universally necessary because many signals and prediction targets change slowly.40:04
Recommender System: Why They Update Models 100 Times a Day
Pushed backGleb Abroskin rejected the description of FunCorp's system as a unified feature store with declarative transformations and one API for offline and online stores.15:00
Tecton 0.6: Notebook-driven Development
ClaimNotebook-driven development can help data engineering, data science, and ML engineering teams improve feature engineering workflows and shorten iteration loops.4:34
MLOps Build or Buy, Startup vs. Enterprise?
ClaimSlack can often add a new recommender quickly by adjusting the features, feature weights, and candidate-fetching method for a use case.16:10
The Future of Feature Stores and Platforms
Pushed backMike Del Balso said feature templates and reusable feature sets are appealing, but customers' differing data and requirements make simple copy-and-paste solutions unlikely to solve most problems.1:06:26
Product Strategy for LLM Features When LLMs Aren't Your Product
Pushed backLLM-based feature engineering should not be viewed only as a prototyping or brainstorming technique; it can be suitable for production in some cases.11:30
AI Innovations: The Power of Feature Platforms
Pushed backNikhil Garg says a feature store and a feature platform are different: a feature store mainly provides storage and serving, while a feature platform also includes computation and other end-to-end capabilities.44:51
RecSys at Spotify
Pushed backA vector database is not always a separate downstream destination for embeddings; in some online use cases it and the feature store serve overlapping roles.38:08
BigQuery Feature Store
Pushed backMalt's BigQuery feature-store approach is not appropriate for every use case because live feature computation, very fresh data, or data too large for memory may require other systems.44:39
Global Feature Store: Optimizing Locally and Scaling Globally at Delivery Hero
Pushed backThe participants disagreed over the definitions of a feature and a feature store before aligning on shared definitions.20:34
Scaling your data and AI from 0-100 with open source
Pushed backThe speakers disputed the assumption that production feature pipelines should be rewritten and maintained by a separate data or ML team after data scientists hand over notebooks.52:54
Hard Learned Lessons from Over a Decade in AI
ClaimTecton's feature store centralized and automated the data pipelines needed to create training sets and provide real-time model features.5:07
Real-time Feature Generation at Lyft
Pushed backA conventional feature store is not sufficient for Lyft's hierarchical, geospatially aggregated features.23:57
Context Engineering 2.0
Pushed backFeatureform's acquisition by Redis does not signal the death of the feature store.0:25
Graph Neural Networks Just Solved Enterprise AI?
Pushed backRelational deep learning is fundamentally different from AutoML that generates many joins, aggregates, features, and model trials.17:34
The Semantic Layer and AI Agents
Pushed backWhether feature stores and semantic layers should be treated as essentially the same type of system.46:09
Real-time features, AI search, Agentic similarities
Pushed backVarant Zanoyan argued that Feather's documentation described the right problems but its implementation had not completed important compute features such as streaming and windowed aggregation.12:01
OpenXData Conference
ClaimAn online feature store provides consistent feature values for training and inference together with discovery, lineage, access control, and versioning.2:02:14