# Arize AI

**Source:** https://geo.sig.ai/brands/arize-ai  
**Vertical:** Developer Tools  
**Subcategory:** AI/ML Observability & Evaluation  
**Tier:** Challenger  
**Website:** arize.com  
**Last Updated:** 2026-04-15

## Summary

Leading AI observability and evaluation platform. Founded 2020, Berkeley CA. $131M total funding. Open-source Phoenix has 2M+ monthly downloads. Private.

## Company Overview

Arize AI is an AI observability and evaluation platform founded in January 2020 by Jason Lopatecki (CEO) and Aparna Dhinakaran (CPO), headquartered in Berkeley, California. It provides tools for monitoring, troubleshooting, and improving AI models in production, covering both traditional ML and LLMs. Key products include the Arize platform and the open-source Phoenix library.

Arize operates a hybrid SaaS plus open-source model. Phoenix has become the most widely adopted AI observability library with 2M+ monthly downloads. Clients include Booking.com, Duolingo, Hyatt, PepsiCo, Uber, and Wayfair. Arize competes with Datadog, Weights & Biases, and LangSmith.

In February 2025, Arize raised $70 million Series C led by Adams Street Partners, with participation from M12 (Microsoft Ventures), Datadog, and PagerDuty, bringing total funding to $131 million.

## Frequently Asked Questions

### What is Arize AI?
An observability and evaluation platform for AI systems, helping teams monitor and improve both traditional ML models and LLMs in production.

### What is Arize Phoenix?
Arize's open-source AI observability library with 2M+ monthly downloads, the most widely adopted tool of its kind.

### How much funding has Arize AI raised?
$131 million total, including a $70M Series C in February 2025 led by Adams Street Partners with Datadog and Microsoft's M12 participating.

### What companies use Arize AI?
Booking.com, Duolingo, Hyatt, PepsiCo, Priceline, TripAdvisor, Uber, Wayfair, and hundreds of other enterprises.

### Who founded Arize AI?
Jason Lopatecki (CEO) and Aparna Dhinakaran (CPO) in January 2020, headquartered in Berkeley, California.

### What is the difference between Arize and Arize Phoenix?
Arize is the commercial enterprise observability platform for monitoring AI models in production — it offers managed infrastructure, compliance features, and enterprise SLAs. Arize Phoenix is the open-source counterpart, a lightweight evaluation and observability library for local development and experimentation. Teams typically use Phoenix during development and Arize in production, though both can be used independently.

### What LLM-specific metrics does Arize monitor?
Arize monitors hallucination rates, toxicity scores, prompt injection attempts, context relevance, groundedness, faithfulness, and response latency for LLM applications. For traditional ML models it tracks data drift, prediction drift, feature importance shifts, and model performance degradation. These LLM-specific evals run continuously in production with configurable alerting thresholds.

### How does Arize integrate with existing ML infrastructure?
Arize integrates via Python SDK, REST API, and native connectors to LangChain, LlamaIndex, OpenAI, Anthropic, and major LLM frameworks. Data can be logged from any cloud environment — AWS, Azure, GCP — and Arize supports OpenTelemetry traces for standardized observability. Enterprise deployments can route data to a private Arize instance within the customer's VPC.

## Tags

b2b, developer-tools, platform, saas

---
*Data from geo.sig.ai Brand Intelligence Database. Updated 2026-04-15.*