# Arthur

> Arthur is a New York-based AI governance and observability company that gives enterprises a single control plane to discover, monitor, evaluate, and govern the AI models and autonomous agents running across their organization. Founded in 2018 by a team out of Capital One and academia, Arthur started with machine-learning monitoring - catching model drift, bias, and performance decay in production - and has since expanded into real-time guardrails, LLM evaluation, and agentic AI governance. Its open-source Arthur Engine and its Agent Discovery & Governance platform help security, compliance, and ML teams ship AI they can trust in regulated industries like finance, insurance, healthcare, and government.

- **Founded:** 2018
- **Headquarters:** New York, New York, United States
- **Founders:** Adam Wenchel (Co-founder & CEO), John Dickerson (Co-founder & Chief Scientist), Liz O'Sullivan (Co-founder), Priscilla Alexander (Co-founder)
- **Team size:** ~39 employees
- **Products:** Arthur (AI Control Plane), Agent Discovery & Governance (ADG), Arthur Engine, Agent Development Toolkit, Policies & Guardrails
- **Notable:** Raised what was reported as the industry's largest-ever ML observability round - $42M Series B in 2022., Open-sourced the first real-time AI evaluation engine (Arthur Engine)., Signed the U.S. Department of Defense as a customer.

## Products & services

- **Arthur (AI Control Plane)** — Enterprise platform to measure, monitor, and govern models and agents across an organization, with SaaS, hybrid, and on-premises deployment.
- **Agent Discovery & Governance (ADG)** — End-to-end platform to automatically discover, inventory, risk-score, and enforce policy across AI agents in production via OTEL, MCP, and cloud APIs.
- **Arthur Engine** — Open-source, real-time AI evaluation engine measuring response relevance, hallucination rate, token counts, latency, PII leakage, and toxicity inside your own environment.
- **Agent Development Toolkit** — Prompt versioning, A/B experiments, tracing, and regression detection covering the full agent lifecycle, built on the open-source Arthur Engine.
- **Policies & Guardrails** — Synchronous, evaluation-based enforcement that blocks prompt injection, sensitive-data leakage, hallucinations, and toxic output in real time.
- **Arthur Bench** — Open-source tool to benchmark and choose the right generative AI model for a company's data and workloads.
- **Arthur Chat** — Retrieval-augmented generation platform to enhance chatbots with a company's own datasets for more accurate, specialized answers.

## Achievements

- Raised what was reported as the industry's largest-ever ML observability round - $42M Series B in 2022.
- Open-sourced the first real-time AI evaluation engine (Arthur Engine).
- Signed the U.S. Department of Defense as a customer.
- Monitored over 1 billion tokens across enterprise deployments in 2025.
- Helped Upsolve catch a critical GPT-5 regression before customer exposure; helped Expel cut ML monitoring time by 50%.
- Launched Agent Discovery & Governance, positioned as a first end-to-end product for managing agentic AI in production.

## Latest updates

- **2025-12** — Launched Agent Discovery & Governance (ADG), an end-to-end platform for discovering, monitoring, and enforcing policy on agentic AI in production.
- **2025-11** — Platform release notes highlight stronger evaluation Engines and a 2026 roadmap focused on policy agents and automated enterprise discovery.
- **2025** — Open-sourced the Arthur Engine, a real-time AI evaluation engine, and introduced the Agent Development Lifecycle (ADLC) methodology.
- **2024-01** — Added support for monitoring AI-powered recommendation systems.

## Links

- Website: https://arthur.ai
- LinkedIn: https://www.linkedin.com/company/arthurai
- Twitter/X: https://twitter.com/itsArthurAI
- GitHub: https://github.com/arthur-ai
- Facebook: https://facebook.com/itsarthurai/

---

Profile page: https://yespress.io/arthur
Published by YesPress — https://yespress.io
Last updated: 2026-07-22