As organizations extend machine learning (ML) capabilities to analyze data generated by increasingly complex applications, New Relic has updated New Relic One’s comprehensive monitoring application to include machine learning processes (MLOps) designed to help manage multiple data and ML models across different business units.

Along with applications, network, infrastructure, browser monitoring, logs and error management, New Relic One is designed to allow data scientists and ML engineers to not only monitor ML model performance but also retrain models after raising alerts, said Jay Figel, general manager of applied intelligence and group vice president for engineering. Products in New Relic.

Observation is a relatively new term in information technology, used to describe the task of monitoring enterprise applications, data flows, and distributed infrastructure. Systems that provide monitoring go beyond former Application Performance Monitoring (APM) software, providing a high-level overview of the IT infrastructure as well as accurate metrics, to allow for effective application, network, data, and security management.

according to Research Report Released by LogDNA for records management applications, 75% of responding companies are still struggling to achieve true observability despite significant investments in tools.

The study, which included 200 top engineering professionals across the United States, showed that two-thirds of organizations currently spend $100,000 or more annually on monitoring tools, with 38% spending $300,000 or more annually.

MLOps helps to monitor the system

The New Relic One update is designed to help alleviate many of the pain points for data scientists, chief among them the changing nature of ML or AI models, as they rely on basic data and code that may become irrelevant as real-world conditions change.

“ML models deteriorate over time,” said Andy Thorai, Vice President of Research and Principal Analyst at Constellation Research. “So you need to monitor the model to measure model performance, skew, model dulling/freshness, model recall, model accuracy and model accuracy metrics. Depending on the application. and usage, models can change in a matter of seconds or can be valid for days/weeks/years in rare cases.”

Feigl said the New Relic One update allows software engineers and data scientists to either import their own data or integrate with data science platforms, as well as monitor machine learning models and interdependence with other application components, including infrastructure.

Currently, New Relic supports data science platforms such as AWS SageMaker, DataRobot, Aporia, Superwise, Comet, DAGsHub, Mona, TruEra, and others.

The company said organizations can create custom dashboards to track the accuracy of machine learning models and generate alerts for unusual changes before they have an impact on business or customers.

Observability to break data silos, speed developers

Another issue facing organizations deploying ML applications, according to New Relic’s Fighel, is how different teams across organizations cannot work with each other efficiently due to different dashboards and separate interfaces.

“There is a huge gap between model producers, AKA data scientists versus model implementers, AKA data engineers, and devops teams. By having tools like these, a model can be produced easily,” Althray said.

The New Relic One platform can help bring teams together even if an organization has already invested in separate data science platforms, by providing a common interface that allows data scientists and other users to import data and view models built on different learning platforms, Feigl said.

This ability can also help treat Sellers InsuranceFeigl said. According to a LogDNA research report, more than half of professionals surveyed said that companies cannot implement the tools they want due to vendor lockouts.

Pricing and availability

New ML capabilities, which are generally available, are offered at no additional cost on the New Relic One platform with a cap of 100 GB per month. However, Feigl said the new system will soon follow the depreciation pricing model.

Some of New Relic’s competitors include Sumo Logic, AppDynamics, Dynatrace, ManageEngine, and the Microsoft Azure Application Insights group.

Copyright © 2021 IDG Communications, Inc.

Source link

Leave a Reply

Your email address will not be published.