Dynatrace’s $915M Arize deal bets AI agents are just another app to monitor
Dynatrace closed its $915 million acquisition of Arize on October 1, joining Arize’s tracing and evaluation of AI agents to The post Dynatrace’s $915M Arize deal bets AI agents are just another app to monitor appeared first on The New Stack .
On October 1, Dynatrace announced the acquisition of AI observability vendor Arize for $915 million. This move integrates Arize's tracing and evaluation of AI agents into Dynatrace's existing monitoring capabilities for applications and infrastructure. As a result, Arize becomes one less independent AI observability vendor on the market, having previously received an investment from Datadog.
Sean O'Dell, a principal product marketing manager at Dynatrace, discussed the deal at the WeAreDevelopers conference in San Jose on September 25, five days before the acquisition closed. O'Dell emphasized that an AI app is still just an application, and when it breaks, the business needs to understand the outage cost while engineers on call must comprehend the cause of the issue.
O'Dell explained that while natural language processing and Retrieval-Augmented Generation (RAG) can be used to evaluate AI apps, the fundamental question remains: "Is it an application?" Teams need to assess whether the AI's response is appropriate, if it's hallucinating, and the financial impact of the AI's actions. He noted that both Dynatrace and Arize already provide answers to these questions.
AI apps, O'Dell pointed out, remain part of a larger ecosystem that includes legacy applications and mainframes. Previously, teams required two separate tools to debug AI-related issues: one for the AI side in Arize and another to investigate software problems through additional traces.
Arize co-founder Aparna Dhinakaran informed The New Stack that teams can debug AI problems within Arize, but software issues necessitate checking separate traces to identify the root cause. Arize also provides Phoenix, a source-available tracing and testing tool specifically for AI apps, along with the community of developers utilizing it. O'Dell highlighted the rarity of such a community, given the early and new nature of AI app development.
Both Dynatrace and Arize possess agents that act upon the data. Dynatrace's Bluebox, introduced in June, analyzes a team's code in GitHub, GitLab, or Bitbucket, compares it to production data, and suggests potential fixes. Arize's Signal performs a similar function for Alyx, Arize's AI assistant, with Dhinakaran revealing that Signal secures approximately 65% to 70% of the pull requests authored by Signal. The remaining third of pull requests still requires human validation.
Written by urgent.news from The New Stack's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
