Best AI Agent Observability, Evaluation, & Governance Companies

Weights & Biases operates as an artificial intelligence (AI) developer platform that provides tools for machine learning and artificial intelligence. The company offers products for building, fine-tuning, and deploying machine learning models, as well as tools for software developers to track and evaluate large language model applications. It serves the AI and machine learning sectors within the technology industry. It was founded in 2017 and is based in San Francisco, California. In March 2025, Weights & Biases was acquired by CoreWeave at a valuation of $1.7B.

Fiddler focuses on artificial intelligence (AI) observability and responsible AI governance in the technology sector. The company provides services such as monitoring, explainable AI, and analytics to support the performance and integrity of machine learning models and large language models. Fiddler serves sectors that need AI governance and risk management, including government and financial services. It was founded in 2018 and is based in Palo Alto, California.

Arize provides tools for AI observability and LLM evaluation within the machine learning and artificial intelligence sectors. The company offers a platform for monitoring, diagnosing, and improving the performance of AI models and applications in production. Arize's tools are based on open-source standards and can integrate with existing AI infrastructure. It was founded in 2020 and is based in Mill Valley, California.

Credo AI offers a platform that automates artificial intelligence (AI) oversight, risk management, and regulatory compliance to facilitate responsible AI adoption. Credo AI's services include AI auditing to ensure system integrity and fairness, as well as educational workshops to empower teams in AI governance practices. It was founded in 2020 and is based in Palo Alto, California.

Patronus AI focuses on automated AI evaluation and security within the AI development sector. The company provides a platform that allows enterprise teams to score the performance of large language models (LLMs), generate adversarial test cases, and benchmark AI systems. Patronus AI serves sectors that require AI evaluation and security measures, including the tech and enterprise AI industries. Patronus AI was formerly known as Zeno AI. It was founded in 2023 and is based in Dublin, California.

AgentOps focuses on creating reliable AI agents within the technology sector. Its main offerings include a suite of developer tools for AI agent development and an observability platform to monitor, test, and analyze AI agents. AgentOps primarily serves clients ranging from startups to large enterprises looking to implement scalable and reliable AI agents. It was founded in 2023 and is based in San Francisco, California.

Aporia provides AI security, reliability, and observability within the technology sector. The company offers guardrails for AI applications, including solutions for prompt injection detection, data leakage prevention, and customizable AI policies. Aporia's offerings are intended for enterprises looking to improve the security and reliability of their AI systems. It was founded in 2019 and is based in Tel Aviv, Israel. In December 2024, Aporia was acquired by Coralogix at a valuation of $50M.

Braintrust is a technology company that builds a platform for developing AI applications within the artificial intelligence sector. The company provides tools for evaluating and managing large linguistic models, including prompt management, performance tracking, and dataset management. Braintrust's solutions include features such as real-time execution trace visualization, monitoring, and the option for self-hosting to meet data control and compliance needs. It was founded in 2023 and is based in San Francisco, California.

Cekura focuses on AI voice agent testing and observability within the technology sector. The company provides services that include scenario generation, persona emulation, and evaluation metrics to assess AI voice agents' performance across various conversational scenarios. Cekura serves the conversational AI industry, offering tools for scenario simulation, monitoring, and performance analysis. Cekura was formerly known as Vocera. It was founded in 2024 and is based in Sunnyvale, California.

Coval focuses on the development of AI agents through automated testing, specifically in chat and voice systems within the AI industry. The company provides simulation and evaluation services that facilitate the creation, testing, and monitoring of AI agents. Coval's solutions are utilized by sectors that develop AI voice and chat agents. It was founded in 2024 and is based in Los Angeles, California.

Deepchecks evaluates and monitors machine learning models, focusing on large language models (LLMs). The company provides solutions for the evaluation of LLM-based applications, aiming to detect and mitigate issues such as hallucinations, incorrect answers, and bias. Deepchecks serve sectors that develop and deploy LLM-based applications, including those involved in content creation processes. It was founded in 2019 and is based in Ramat Gan, Israel.

Langfuse focuses on LLM engineering. It provides tools for observability and improvement of LLM applications. The company offers services including metrics, evaluations, prompt management, and a playground to debug and enhance LLM apps. Langfuse is designed to work with any model or framework. It was founded in 2022 and is based in Berlin, Germany.

Larridin operates within the business technology sector, providing services that enhance organizational productivity. The company's services include insights into workforce optimization, employee engagement, and resource allocation efficiency, as well as analyzing the productivity of AI agents. Larridin serves sectors that require productivity and performance analytics, such as finance, human resources, technology, and sales. It was founded in 2024 and is based in San Francisco, California.

Traceloop specializes in observability solutions for large language model (LLM) applications within the technology sector. The company offers monitoring services that provide real-time alerts, insights, and tools for backtesting, debugging, and gradual rollout of changes to LLM applications. Traceloop primarily serves engineers and developers in the tech industry who require robust monitoring and testing solutions for their LLMs. It was founded in 2022 and is based in Tel Aviv, Israel.

Vijil focuses on enhancing the trustworthiness of autonomous agents within the AI software industry. Its main offerings include private cloud services that enable AI engineers to measure, improve, and maintain trust in AI agents, ensuring they are open, safe, and secure. Its services are designed to harden large language models during development, defend them during operation, and continuously evaluate their reliability, security, and safety. It was founded in 2023 and is based in Menlo Park, California.