Welcome to the feature store newsletter brought to you by Logical Clocks, where we in a monthly cadence highlight the latest news, events, and insights as we help companies become successful in their machine learning transformation journey and empower businesses to be applied-AI model-driven companies.
How can a Feature Store improve collaboration between data scientists and data engineers?
One of the key benefits of a feature store is its ability to improve collaboration between data engineers, who transform backend data into features, and data scientists, who use the features to train models. The feature store acts as an API, defining where feature engineering pipelines stop and training pipelines start. In classical engineering terms, this enables data engineers and data scientists to work independently towards the common goal of end-to-end machine learning pipelines.
This is the highlighted topic for this month and we share with you an article published in VentureBeat that discusses why your teams should be working together and how to foster a more collaborative development environment between data scientists and data engineers.
Read here
Editor’s Picks
The Sequence Edge#10: Feature Extraction
The 10th issue of The Sequence Edge newsletter explains the difference between feature extraction and feature selection, explores a feature visualization method known as Activation Atlases, and reviews the Hopsworks feature store platform.
Read here
The Essential Architectures for every Data Scientist and Big Data Engineer
Sharmistha Chatterjee from Publicis Sapient does a comprehensive list of feature store architectures for data scientists and big data professionals.
Read here
Butterfree: A Spark-based Framework for Feature Store Building
Butterfree is a new open-source Spark-based framework for feature store creation with S3 and Cassandra for offline and online features.
Read here
Drive a Culture of Experimentation For Effective AI/ML
Splice Machine’s CEO, Monte Zweben, wrote an article on Medium about how feature stores can help drive a culture of experimentation in AI/ML teams.
Read here
Stimulate ML Development using a Feature Store
Jayesh Patel of Rockstar Games wrote a blog, ahead of a talk at the Data Innovation Summit in August 2020, about their motivation for using a feature store for ML.
Read here
Common problems taking ML from lab to production
Charna Parkey from Kaskada takes a look into disappearing data and degraded performance preventing ML models from shipping.
Read here
Upcoming Webinars & Events
Attend this webinar to learn how to use deep learning (both supervised & unsupervised learning) and graph embeddings as potential technologies to increase anomaly detection rates and reduce costs associated with fraud and anti-money laundering, using our Hopsworks data platform.
Register here
Latest Videos
Time Travel and Provenance for Machine Learning Pipelines
Watch this video to learn what "implicit provenance" for ML is, how Alex Ormenisan from Logical Clocks added support in, and how it helps with debugging and governing ML pipelines.
Watch here
Manage your own Feature Store on Databricks with Hopsworks
Watch this demo to learn how to engineer your features on Databricks and publish them to Hopsworks Feature Store.
Watch here
Job Opportunities
Sr Software Engineer – ML Feature Store
Twitter is hiring a software engineer to join their ML Feature Management (MLFM) team in the USA.
Apply here
ML Engineering Manager
Wildlife Studios is hiring a ML engineer to structure their feature store to architect their AI microservices framework.
Apply here
Data Management Consultant
The Granite Solutions Group is looking for a new consultant to design a feature store platform and feature governance strategy.
Apply here
Software Engineer, ML Serving Platform
Pinterest is hiring a new addition to their ML Serving Platform team who will help to build a feature store.
