Arize

Arize AI is an observability and evaluation platform for machine learning models and AI applications, founded in January 2020 in Berkeley, California by Jason Lopatecki and Aparna Dhinakaran. It started as a traditional ML monitoring tool covering drift detection and model performance dashboards, and has since expanded into LLM and AI agent observability. The company raised $131 million in total funding, including a $70 million Series C in February 2025 led by Adams Street Partners, with Datadog and PagerDuty participating.

The platform has two products. Arize AX is the commercial platform with session- and span-level tracing, LLM-as-a-judge evaluations, real-time alerts through PagerDuty and Slack, drift detection, and an AI debugging assistant called Alyx, carrying SOC 2, GDPR, and HIPAA compliance. Arize Phoenix is the open-source side built on OpenTelemetry, handling tracing, evaluation, prompt management, and experimentation, which crossed 10,000 GitHub stars in June 2026 and integrates with LangChain, LlamaIndex, the OpenAI Agents SDK, and CrewAI.

Limitation: in August 2026 Dynatrace signed a definitive agreement to acquire Arize, expected to close later that quarter subject to regulatory review. Teams standardizing on Arize should check the new owner's announcements for license, hosting, and maintenance commitments before committing long-term.

Sign up for Arize AX cloud or self-host Phoenix via Docker; instrument with OpenTelemetry or the OpenInference standard.