Starting a company can involve far more work than simply coming up with a business idea.
Founders may need to register a company, create payment systems, set up email accounts, buy cloud infrastructure, manage phone numbers and connect a growing collection of software tools.
Naïve, a startup building infrastructure for AI agents, believes much of that work could eventually be automated.
The company has raised $28.5 million in a Series A funding round led by Nexus Venture Partners as it expands a platform designed to help AI agents handle many of the tasks involved in setting up and running a business.
Naïve provides developers with infrastructure that can connect AI agents to services required for business operations.
Its technology can help automate processes including company formation, payments, email, phone systems, databases, cloud computing and connections to services such as Stripe and QuickBooks.
The idea is part of a rapidly growing movement around autonomous businesses.
Instead of humans manually completing every task, AI agents could eventually be given the ability to carry out specific actions on behalf of a company.
A founder might provide instructions, set limits and approve sensitive decisions while AI systems handle a large amount of the underlying administrative work.
Naïve says developers can connect its infrastructure to tools such as Cursor, Claude Code and Codex, allowing AI agents to interact with the systems required to establish a business.
The company still requires humans to complete certain processes involving identity verification and payments.
But much of the surrounding work can potentially be automated.
Naïve says more than 30,000 developers have signed up to use its technology within months of its launch.
That early interest highlights the growing demand for AI systems capable of doing more than generating text.
Businesses are increasingly interested in software that can take action.
An AI agent that can write a marketing email may be useful.
An AI agent that can also create an account, connect software, process information and carry out a series of tasks could have a much larger impact.
However, giving AI agents greater autonomy also creates new risks.
A system capable of accessing payment platforms, cloud infrastructure and business data needs to operate within strict controls.
Naïve is attempting to address that issue through a governance layer that allows users to set budgets, restrict the actions available to agents and require human approval before sensitive decisions are made.
The startup is also investing in infrastructure designed to reduce the cost of running large numbers of AI agents.
That could become a major challenge as autonomous software becomes more common.
AI agents consume computing resources whenever they make model requests, process information or remain active while waiting for the next task.
For a business operating hundreds or thousands of agents, those costs could quickly become significant.
Naïve is developing systems for model routing, memory and agent orchestration in an effort to make those operations more efficient.
The company is also building a serverless environment that allows agents to run without requiring a full virtual machine for every process.
The funding will be used to expand those projects and hire additional researchers.
Naïve’s vision may still sound ambitious.
But the company is part of a growing group of startups asking a fundamental question about the future of work: what happens when software is capable of helping create and operate the businesses that use it?
The answer may not be a completely autonomous company anytime soon.
But if startups such as Naïve succeed, building one could eventually involve far fewer manual steps than it does today.
