Naïve Raises $28.5M for AI Company Infrastructure

A significant funding round has propelled Naïve, a startup focused on building the operational backbone for autonomous businesses, into its next growth phase. The company announced a $28.5 million Series A investment, designed to accelerate the development of its unified platform that empowers AI agents to manage and execute business functions. This substantial capital infusion was spearheaded by Nexus Venture Partners, with crucial backing from prominent venture capital firms including Y Combinator, Zetta Venture Partners, and Liquid 2 Ventures. A notable group of angel investors also participated, featuring industry heavyweights such as Gokul Rajaram, Apollo co-founder Tim Zheng, former HubSpot executive JD Sherman, Amazon executive Gert Lanckriet, DocuSign executive Robert Chatwani, and Codecademy co-founder Zachary Sims.

Founded by a pair of young entrepreneurs, Sean Dorje and Dennis Zax, both 20-year-old University of California, Berkeley dropouts with a history of collaboration since their early teens, Naïve addresses a critical gap in the burgeoning field of AI-driven operations. Their prior success in building and divesting the machine-learning company ezML before their university careers underscores their precocious talent. The core of Naïve's offering is an operating stack engineered to equip AI agents with the technical, financial, and organizational capabilities necessary to function as independent business entities. This moves beyond mere code generation to encompass the complexities of real-world company management.

The current challenge Naïve aims to solve lies in the fragmented nature of services required to transition AI-generated software into a fully operational business. While AI agents can rapidly produce functional code, integrating essential services like business incorporation, payment processing, communication tools, cloud infrastructure, and accounting systems remains a significant hurdle for developers. Many of these existing systems were designed with human users in mind, creating integration complexities and governance challenges for autonomous agents. Naïve's solution centralizes these disparate functions through a single configuration file and a unified API, simplifying the process of provisioning and managing operational infrastructure.

Naïve's platform provides a comprehensive suite of tools for autonomous company formation and operation. This includes services for business incorporation, virtual payment cards, dedicated email inboxes, and virtual phone numbers, enabling AI agents to operate through identifiable legal and economic entities. The cloud services component offers essential infrastructure such as relational databases, hosting, computing resources, and authentication. Furthermore, Naïve introduces innovative serverless environments for AI agents, representing them as lightweight, serializable states rather than continuously provisioned virtual machines. This architectural choice, utilizing V8 isolates, allows for rapid cold starts—reportedly as low as 2.3 milliseconds—and charges based on active execution periods, significantly enhancing cost efficiency.

Beyond infrastructure, Naïve places a strong emphasis on governance and control. The platform incorporates a governance gateway that scrutinizes agent actions before execution, allowing developers to set budgets, define approval workflows, and establish capability restrictions. This ensures that AI agents operate within defined parameters, mitigating risks associated with autonomous decision-making. The company is also investing heavily in optimizing AI inference and agent orchestration. Their strategy involves routing requests to the most cost-effective AI models capable of meeting task requirements and developing a shared memory layer that distills information into structured facts, aiming for superior recall with reduced token consumption. This focus on efficiency is critical as AI agent spend is projected to reach trillions globally.

The newly acquired capital will be strategically allocated to four key research and development areas: serverless runtimes, inference optimization, shared memory systems, and multi-agent orchestration. Naïve's approach to serverless runtimes, for instance, promises substantial cost savings compared to traditional continuously available virtual machines, with reported per-agent resource requirements of approximately 1.2 megabytes. Their inference research aims to reduce model expenses by employing smaller language models for routine tasks and batched inference for more complex processes. The company's ambition is to make autonomous companies economically viable and scalable, ensuring that human oversight remains central to critical financial and operational decisions, thereby maximizing the useful work performed by AI agents for every token consumed.