Arize
Mill Valley, US · Delaware corporation · 233 employees on LinkedIn · 14 known investors
Arize provides observability and evaluation tools for AI systems and large language models, enabling development teams to monitor, debug, and improve AI agents in production.
Also known as Arize AI · Arize AI, Inc.
Founders & leadership
Board


Investors · 14
Also in the syndicate · 2
Reported raises · per SEC filings
Form D private placements$42M disclosed across 2 of 4 rounds · 2020–2025
▶$69.1MofferedJan 2025 · 23 investors · Other TechnologyRule 506(b)
- APARNA DHINKARANDirector
- DHARMESH THAKKERDirector
- BRETT WILSONDirector
- JASON LOPATECKIExecutive Officer, Director
- MORGAN GERLAKDirector
- ASHU GARGDirector
- Offering amount
- $69.1M
- First sale
- Jan 2025
- Incorporated
- Corporation, Delaware
- Federal exemptions
- 06b
▶$38MraisedOct 2022 · 11 investors · Other TechnologyRule 506(b)
- Ashu GargDirector
- Morgan GerlakDirector
- Aparna DhinakaranDirector
- Dharmesh ThakkerDirector
- Jason LopateckiExecutive Officer, Director
- Brett WilsonDirector
- Offering amount
- $38M
- Amount sold
- $38M
- First sale
- Aug 2022
- Incorporated
- Corporation, Delaware, 2020
- Federal exemptions
- 06b
▶$4MraisedFeb 2020 · 6 investors · Other TechnologyRule 506(b)
- Ashu GargDirector
- Jason LopateckiExecutive Officer, Director
- Brett WilsonDirector
- Offering amount
- $4M
- Amount sold
- $4M
- First sale
- Feb 2020
- Incorporated
- Corporation, Delaware, 2020
- Federal exemptions
- 06b
Source: SEC EDGAR Form D. Amounts as filed; amended filings shown once at their latest values.
Company profile
researched Aug 2026Arize AI provides an AI engineering platform for observing, evaluating and improving AI agents and applications. Its commercial product, Arize AX, captures traces of agent runs (inputs, outputs, tool calls, latency, token usage and cost), surfaces recurring failures as ranked issues through a feature called Signal, scores quality using LLM-as-a-judge, agent-as-a-judge and code evaluators, and validates changes through datasets and experiments before deployment. The platform includes Alyx, an in-product AI engineering agent that runs evaluations, debugs issues and proposes fixes, and ADB, a datastore for GenAI traces that stores data in open formats for connection to BigQuery, Databricks or Snowflake. Coverage spans generative AI, AI agents, traditional machine learning and computer vision, including chatbots, RAG systems and copilots.
Alongside the commercial product, Arize maintains Phoenix, an open-source, local-first platform for tracing, evaluation, experimentation and prompt iteration, launched in 2023. The company also originated OpenInference, a set of GenAI semantic conventions built on OpenTelemetry. Arize AI Copilot, described as an AI assistant for AI engineers with over 50 built-in skills, launched in 2024. The company states its platform integrates with 40 or more models, frameworks and AI tools, including OpenAI, Anthropic, Google, Amazon Bedrock, LangGraph, LangChain, LlamaIndex, CrewAI, the OpenAI Agents SDK and DSPy, and runs across Google Cloud, AWS, Azure and self-hosted environments.
Founding story
Arize was founded in 2020 by Jason Lopatecki and Aparna Dhinakaran, who shared a frustration that machine learning investments were growing rapidly while teams lacked the tools to understand why models behaved as they did once deployed. The founders' initial seed deck opened with the line "We Make AI Work," and the company set out to give teams the means to troubleshoot and improve AI performance in production.
Business model
Arize sells a managed SaaS and self-hosted platform (Arize AX) on a tiered subscription basis while distributing an open-source tool (Phoenix) that developers can host locally, with an upgrade path to the commercial product. A dedicated startup pricing program is offered.
Tiered subscriptions: AX Free at $0 per month (25k spans, 1 GB ingestion, 15-day retention, 10 Signal issues per month, SaaS only), AX Pro at $50 per month (50k spans, 10 GB ingestion, 30-day retention, 25 Signal issues per month), and AX Enterprise with custom pricing, custom volumes and retention, unlimited Signal issues, SaaS or self-hosted deployment, enterprise SSO, audit logs and SLAs. Users and evaluations are unlimited on all tiers.
Traction
The company's site reports 1 trillion spans processed, 1 billion evaluations per year and 5 million downloads per month; the February 2025 blog post cited over two million monthly downloads for Phoenix. The about page lists a large employee roster and 28 open positions.
Latest developments
In February 2025 Arize announced a $70 million Series C led by Adams Street Partners and an expanded Microsoft partnership covering Azure AI Studio and Azure AI Foundry integrations. Its current product line centers on Arize AX with Signal, the Alyx AI engineering agent, the ADB trace datastore and agent-native debugging workflows for coding agents. The homepage announces "A new chapter begins with Dynatrace."
▸Full profile — market position, technology, go-to-market, geography, history, risks & controversies
Market position
Arize positions itself as a framework-independent evaluation and observability suite in the AI observability category, describing its February 2025 Series C as the largest-ever investment in AI observability and Phoenix as a widely adopted open-source AI observability library.
Arize emphasizes framework and vendor independence, open standards (OpenInference on OpenTelemetry) rather than proprietary trace formats, a unified platform spanning development and production and combining evaluation with observability, flexible SaaS or self-hosted deployment with US, EU or CA data regions, and an accompanying open-source offering in Phoenix.
Technology
The platform is built on OpenTelemetry and OpenInference semantic conventions, avoiding a proprietary trace format. Components include distributed tracing with span, trace and session-level evaluation; online evaluations on traces and spans and offline evaluations on datasets; agent trajectory visualizations; custom dashboards, metrics and monitors; prompt playgrounds, versioning and prompt-learning optimization; human annotation and labeling queues; the Alyx in-product agent; Signal for automated issue clustering; and ADB, an observability datastore using open storage formats. Agent-native workflows, skills and MCP servers support coding agents such as Cursor, Claude Code and OpenCode.
Go-to-market
Product-led entry through a free tier, a self-serve $50/month plan and a terminal/CLI install path, combined with open-source distribution of Phoenix, developer documentation, a Slack community, newsletter and cookbooks, and an enterprise sales motion with demos, dedicated support, onboarding kickoffs and a customer advisory board. Cloud and ecosystem partnerships include Microsoft Azure AI Studio and Azure AI Foundry, Google Cloud and NVIDIA AI microservices.
AI engineers, machine learning practitioners and enterprise AI teams building and operating LLM applications, RAG systems, copilots and multi-agent systems; tiers address individual builders, AI-native teams, startups and large enterprises with self-hosting and compliance requirements. Named users referenced on the site include Wayfair, PagerDuty and TripAdvisor.
Geography
Headquartered in Mill Valley, California. The platform offers US, EU and Canada data regions and supports Google Cloud, AWS, Azure and self-hosted deployments; staff roles include EMEA-focused solutions and sales positions.
History
Founded in 2020 as a machine learning observability platform for detecting and troubleshooting model issues in production. The company filed with the SEC to raise $38 million in equity (reported in a 2022 item). It launched the open-source Phoenix tracing and observability tool in 2023, and Arize AI Copilot for AI engineers in 2024. In February 2025 it announced a $70 million Series C and expanded its Microsoft partnership around Azure AI Studio and Azure AI Foundry. The company's homepage subsequently referenced "A new chapter begins with Dynatrace."
Risks & controversies
The company operates in a competitive AI observability market alongside investors and partners that also offer monitoring products, such as Datadog and PagerDuty, both of which participated in its Series C.
Compiled by commissioned research from 8 cited public sources — announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed — not independently audited.
Founder mafia
2 people who came through Arize went on to found or lead other companies.
Competitors · 4
by search overlapCompanies competing with Arize for the same Google search keywords, organic and paid, via search-intersection analysis.
Timeline · 5
launches, deals, and filingsArize announced a $70 million Series C led by Adams Street Partners, with participation from M12, Sinewave Ventures, OMERS Ventures, Datadog, PagerDuty, Industry Ventures and Archerman Capital, plus existing investors Foundation Capital, Battery Ventures, TCV and Swift VC. The company described it as the largest-ever investment in AI observability.
$70M source ↗
Arize said it was expanding its partnership with Microsoft, adding deeper integrations with Azure AI Studio and the Azure AI Foundry portal, SDK and CLI, while continuing technical integrations with Google Cloud and NVIDIA AI microservices.
Arize launched an AI assistant for AI engineers with over 50 built-in skills, covering tasks such as debugging agent traces, writing evaluations and optimizing prompts.
Arize launched Phoenix, an open-source AI observability and performance tracing tool for local or self-hosted use.
Filings with the U.S. Securities and Exchange Commission indicated Mill Valley-based Arize AI was raising $38,000,000 in new equity funding, with CEO Jason Lopatecki cited as leading the effort.
$38M source ↗
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
Legal entities · 1
corporate structureIn the news
▸Research sources · 8
primary sources listed
- Arizearize.com · web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Arize do?
- Arize AI builds an observability and evaluation platform for AI agents, LLM applications and machine learning models.
- Who founded Arize?
- Arize was founded by APARNA DHINKARAN, Jason Lopatecki, Aparna Dhinakaran.
- Who are Arize's investors?
- Arize's investors include Battery Ventures, Collective Liquidity, Evolution Equity Partners, Foundation Capital, M12 (Microsoft's Venture Fund), Nimble Partners, OMERS Ventures, Runa Capital and 4 more.
- How much funding has Arize raised?
- Arize has disclosed $42M raised across 2 of its 4 known rounds.
- Where is Arize headquartered?
- Arize is headquartered in Mill Valley, US.







