Sapiom, a company developing foundational infrastructure for AI agents, has successfully closed a $35 million Series A funding round led by Dragonfly. This new investment brings the company's total capital raised to $50 million, following a $15 million seed round just six months prior. The funding will accelerate Sapiom's mission to build the operating system for a new generation of autonomous software operators.
The Challenge of Autonomous AI
The rapid advancement of AI has introduced a new kind of operator: the autonomous agent. However, most of the world's digital infrastructure was designed with human users in mind, creating significant barriers for software that acts independently. Systems for payments, credentials, and security often treat autonomous agents as intruders, limiting their potential and creating operational friction.
Transitioning AI agents from controlled demonstrations to reliable, real-world production environments presents immense challenges. Businesses struggle with ensuring continuous operation, managing costs, controlling agent actions, and recovering from failures at scale. These escalating costs and operational complexities are expected to lead to the cancellation of a significant number of agentic AI projects.
Sapiom's Production-Ready Infrastructure
Sapiom addresses these issues by providing a production-grade infrastructure layer that sits between an agent and the digital services it interacts with. The platform enforces predefined rules and permissions before an agent acts, ensuring its behavior remains bounded, visible, and secure. This allows agents to perform tasks autonomously without requiring human intervention or overly broad credentials.
The company's platform is delivered through three core products: Sapiom Router, Agent Studio, and Runtime. The Router optimizes AI workloads by selecting the most cost-effective models, while the Agent Studio provides tools to build and deploy agents. Sapiom Runtime offers a fully managed environment for operating agents in production, handling everything from security sandboxes to automated recovery protocols.
The platform's effectiveness is highlighted by significant customer results, including helping one client reduce its monthly AI inference bill from $1.2 million to just $100,000. By providing granular visibility into agent activity, Sapiom enabled the company to understand and optimize its costs. The platform has already processed over 270 million transactions and currently powers more than 100,000 agent runs each day.
Strong Investor Backing and Future Vision
The Series A round was led by Dragonfly, with Haseeb Qureshi, the firm's Managing Partner, joining Sapiom's board. The round also saw participation from Accel, Gradient, and Coinbase Ventures, along with continued support from existing investors like Menlo Ventures and Anthropic. This strong backing from prominent venture capital firms validates Sapiom's approach to solving a critical infrastructure gap in the AI ecosystem.
With this new capital, Sapiom plans to enhance its platform and solidify its position as the standard for building and governing autonomous agents. The company's vision extends beyond cost management to removing all barriers that prevent developers from deploying AI agents at scale. The goal is to provide the essential infrastructure needed to support what they anticipate will be the next trillion agents.
As businesses increasingly deploy autonomous AI, the need for a robust, secure, and efficient operating system becomes paramount. Sapiom's successful funding round underscores the market's recognition of this critical need for specialized infrastructure. The company is now well-positioned to provide the foundational layer that will enable the next wave of innovation in autonomous software.