On January 23, 2023, Arize AI, an industry leader in machine learning observability, introduced a connection solution for BigQuery, Redshift, Delta Lake, and Snowflake data lakes. As a result, Arize customers with centrally managed inference stores can rapidly connect their ML table data to Arize via Data Lake Connectors for strong model observability.
Arize is the market leader in terms of both volumes of models and predictions tracked, surpassing billions of predictions daily. So far, ML observability platforms have struggled to simplify their deployments while simultaneously managing billions of predictions and complicated monitoring services, such as embedding drift. The latest Arize version for data connectors expands clients' connectivity options with the widely used data lakes.
The launch comes as the machine learning community converges on various MLOps designs. A modern machine learning architecture approach involves storing inference data in a data lake. ML teams are designing these ML data lakes to power feature stores for feature serving and an inference store for analysis.
Arize Data Lake Connectors are intended for smooth integration with today's data lake architectures. Among the benefits of connecting directly to the ML data store are the following:
Financial savings can be substantial as compared to alternative techniques of ML monitoring.
Teams can run off of an SSOT
Integration and onboarding are accelerated and simplified
Arize now connects with cloud storage providers (such as Google Cloud Platform, Microsoft Azure, and Amazon Web Services), Python pipelines through an SDK, and Kafka Streaming. With this release, users of data lakes may get real-time model insights more efficiently than ever before. In addition, the platform provides fully managed built-in connectors as part of its cloud and VPC platform, eliminating the need for users to build and maintain complex data pipelines or use a separate ETL tool and providing real-time model performance analysis and monitoring.
About Arize AI
Arize AI offers an ML observability platform that monitors models and provides insights for troubleshooting production AI. The firm was established in 2020 in Berkeley, California. It allows corporate clients to monitor the performance of AI models using software that searches for unanticipated biases in algorithms, does root-cause analysis when issues arise, and enhances overall model performance. Arize operates in the background, analyzing internal and external data to forecast demand and eliminate supply chain network errors and sales losses. Adobe, Chick-fil-A, and eBay are among the fifty clients of Arize AI.