10.15.2021

New Releases, Enhancements, Changes + Arize in the News!

What's New

Performance Troubleshooting

An opinionated flow to showcase and search for time ranges and dips in data. Users can now filter by cohorts of data and interact with the graph in more nuanced ways. While the Arize platform defaults to your particular model’s preferred performance metric, we are able to support an array of different performance metrics.

Enhancements

Fully Customizable Dashboards

Fully re-vamped dashboard creation and editing flow with flexibility, customizability, a11y, and usability in mind. With the re-vamped dashboards, you can fully customize your layout with a simple click-and-drag. Widgets are automatically placed in an intuitive 3-column layout but you can adjust your layout to your heart’s desire. This has been a long-requested feature and surely a welcome addition for our existing customers.

Performance Troubleshooting - Dataset B

New support for a comparison dataset! Now, in addition to rich filtering capabilities, users can also troubleshoot performance degradation by comparing it to another dataset. What's also included:
  • A troubleshoot button from a performance monitor that drops you into the performance tab to drill down on the issue
  • Creation of a dataset with environment batches support
Support for searching for models by name in the Models landing page enabling our customers with large numbers of models to quickly find what they're looking for!

In the News

Arize AI Raises $19 Million Series A As Organizations Move To Address ML Observability, the Missing Foundational Piece of ML infrastructure

We’re growing! We officially announced our $19M Series A round of financing led by Battery Ventures in response to the growing demand for our observability solutions amongst organizations that rely on ML models for critical business applications. Read More
Introducing Amber Roberts, Arize's Newest ML Sales Engineer
As an astrophysicist and data scientist, she helped answer questions about the universe. Now, she’s here to help implement your ML observability strategy. Meet Arize’s newest ML sales engineer, Amber Roberts Read more​
​

The Rise of ML Observability Webinar

Arize AI's first public demo with co-founders Aparna Dhinakaran and Jason Lopatecki! An overview of ML Observability since the conception of Arize AI in 2020. Learn how to detect, diagnose, and resolve ML model performance issues with Arize AI. Watch it​