Professional networking site owned by Microsoft Corp LinkedIn Donates another project to the open source community.
Today she said she is handing over control of Feathr . Features Store For machine learning projects to The Linux Foundation’s Artificial Intelligence and Data Foundation.
feathers It is a tool developed by LinkedIn and first introduced in 2017. It is designed to make serving features in machine learning easy, faster and more scalable, especially for real-time AI applications. LinkedIn engineers Hangfei Lin and Jinghui Mo said in a blog post that the company’s AI teams are using Feathr to store, transform, render, and share features with low latency and high throughput.
More specifically, Feathr acts as an abstraction layer between raw data and machine learning models that helps standardize and simplify defining, transforming, presenting, storing, and accessing features within machine learning workflows and applications. Developers can then focus more on engineering the features, letting Feathr take care of the data sequence formats by connecting to different databases. It also provides performance improvement and credential management capabilities.
With Feathr, machine learning developers can only define features once and use them in multiple scenarios, such as model training and model servicing. They can also connect to various offline data sources, such as data lakes and data warehouses, and turn the data within them into machine learning features.
Since its launch, Feathr has come to support many of LinkedIn’s AI applications, as it is used to manage thousands of features. According to Lin and Mo, it has helped teams reduce the time it takes to add and try new features from weeks to days, with performance up to 50% faster than the custom feature processing pipelines it replaced.
Lin and Mo said that since doing so, Feathr has achieved rapid adoption among the machine learning process community and is now used in businesses of various sizes, across many industries. Moreover, these people not only use the software, but also help contribute to its development. “It’s clear that many others suffer from the same pain points that Phaser aims to address,” Lin Wu Mo said.
By donating Feathr to LF AI & Data, it is hoped that Feathr can grow and develop faster, increasing its visibility, user base and shareholder base. At the same time, Feathr’s core development team expects to gain more opportunities to collaborate with other companies and projects that use the software.
One goal is to achieve richer support for online stores through integration with Milvus and JanusGraph, for example. In addition, Feathr’s proponents want OpenLineage’s open data genealogy standard to be adopted.
“We are excited to welcome Feathr to LF AI & Data and to be part of our portfolio of technical projects with over 16,000 developers,” said Dr. Ibrahim Haddad, CEO of LF AI & Data. “We aim to support Feathr to expand its user base, grow its developer community, become a leader within its own category, and enable opportunities for collaboration and integration with other projects.”