Agency AI, the creators of the industry-leading AgentOps.ai platform, raised $2.6 million in pre-seed funding led by 645 Ventures and Afore Capital.
The new funds will enable the company to further develop and scale its AI Agent observability and testing platform, ensuring reliability and safety for enterprises deploying AI Agents.
Led by Alex Reibman, Adam Silverman, and Shawn Qiu, the platform is now used by thousands of teams, monthly. AgentOps.ai integrates seamlessly with leading AI Agent frameworks, including AutoGen from Microsoft Research, CrewAI, LlamaIndex, and MultiOn, providing unparalleled support and compliance for AI Agent developers.
What started as an internal tool to manage consulting projects on implementing a variety of AI Agents for Fortune 500 corporations quickly turned into full-fledged startup.
“We make the feedback cycle a lot quicker,” explained Silverman in an interview with StartupHub.ai.
AgentOps provides developers and enterprises with the tools needed to monitor, debug, and optimize their AI Agents on a granular level. By logging every interaction and decision made by an Agent, AgentOps offers full visibility into the agent's performance, making it easier to identify and address issues. The platform also helps reduce operational costs by optimizing Agent sessions and caching data.
In addition to its existing integrations, Agency recently announced native support for Cohere’s foundation model, Command R+.
“Our platform is designed to be flexible and robust, accommodating the diverse needs of AI developers," asserted Silverman. He sees Agency AI analogous to MDM (Mobile Device Management)–tracking every single digital movement–for AI agents, providing the necessary tools to monitor, debug, and optimize Agents, in real-time.
Silverman, at just 26 years old, brings a wealth of entrepreneurial experience, starting in high school to founding a TikTok marketing agency in university that was later acquired by a public company.
“In the next year, you will see enterprises employing thousands of Agents in everything from sales to research," predicted Silverman. "AgentOps is the central place where all the data is managed. Whether you're buying an Agent off-the-shelf or having an engineering team build one, there will always be an audit trail within AgentOps. In large enterprises, it is important to know what your employees do. Agents represent new members of the workforce and require the same auditability and accountability that exists with employees.”
As for the development pace of Agentic AI, Silverman believes the industry is still in a phase akin to baseball's 'spring training'–the game hasn't even started.
