1. [Home](https://www.g2.com/)
2. ...
3. [AI Agent Observability Software](https://www.g2.com/categories/ai-agent-observability)
4. [Arize AX Discussions](https://www.g2.com/products/arize-ax/discuss)

[
 ![Product Avatar Image](https://images.g2crowd.com/uploads/product/image/large_detail/large_detail_884a506c81897364f1c3d94fd3c746c4/arize-ax.png "Product Avatar Image")
](/products/arize-ax/reviews)

[

Arize AX

](/products/arize-ax/reviews)

(77)4.4/5

Arize AI is the continual learning and AI engineering platform for observing, evaluating, and improving AI agents and LLM applications across development and production. Trusted by leading AI startups, 25% of Fortune 100 companies, and 150+ enterprises, including Uber, DoorDash, Reddit, and Atlassian. Traditional APM tells teams whether an application is fast and available. Arize goes further, showing whether an AI system behaved as intended and delivered a high-quality, trustworthy response. OBSERVE how your agents actually behave. Trace every step of a session, including prompts, tool calls, retrievals, chains, and multi-agent swarms. ADB, Arize’s purpose-built datastore, unifies traces and eval data in open formats. Its elastic architecture supports real-time streaming and high-volume querying while allowing teams to access their data from existing tools and warehouses without exporting or duplicating it. EVALUATE agent quality using those same traces. Run LLM-as-a-judge, code-based, and Agent-as-a-Judge evaluations to assess quality and score outcomes. Run offline evaluations on datasets to test and compare changes before release, and online evaluations on production traces to monitor quality and detect regressions over time. IMPROVE CONTINUOUSLY by turning production feedback into an agent improvement loop. Signal, Arize’s always-on Agent SRE, continuously reviews production traces to surface emerging issues and failure patterns. Connect to a repository and Arize managed agents can investigate issues and propose fixes as pull requests for human review. Engineers can also use Arize Skills to investigate traces, create datasets and evals, and run experiments from Cursor, Claude Code, Codex, and other coding agents. Arize AI, the team behind OpenInference, provides open, portable instrumentation built on OpenTelemetry, with integrations across more than 40+ models, frameworks, and tools. Arize meets production-grade security and compliance requirements, including SOC 2 Type II, ISO 27001, HIPAA, and GDPR. The team also maintains Phoenix, the open-source AI observability and evaluation platform used by AI engineers worldwide.

Show More

When users leave Arize AX reviews, G2 also collects common questions about the day-to-day use of Arize AX. These questions are then answered by our community of 850k professionals. Submit your question below and join in on the G2 Discussion.

* * *

### 48.0

Nps Score

### All Arize AX Discussions

Search

Most CommentedMost HelpfulCommon QuestionsNewest

All DiscussionsDiscussions with CommentsCommon QuestionsDiscussions without Comments

FilterFilter

Filter byExpand/Collapse 

Sort by

Most Commented

Most Helpful

Common Questions

Newest

Filter by

All Discussions

Discussions with Comments

Common Questions

Discussions without Comments

Sorry...

There are no questions about Arize AX yet.

## Start a New Software Discussion

Have a software question?

Get answers from real users and experts

Start A Discussion

* * *

 ![Product Avatar Image](https://images.g2crowd.com/uploads/product/hd_favicon/06b0929a51b887a0ce5b33ca2ef8ca20/arize-ax.svg "Product Avatar Image")

### Have you used Arize AX before?

Answer a few questions to help the Arize AX community

[
Yes
](javascript:void(0))[
Yes
](https://www.g2.com/login?context=product_review&return_to=https%3A%2F%2Fwww.g2.com%2Fproducts%2Farize-ax%2Fdiscuss%3Fsmall_ask%3Darize-ax)
No