Naïve lands $28.5M as 30,000 devs sign up to let AI agents run entire businesses
Curated by the Inblix editorial team
The developer exodus from drudgery now extends beyond writing code to incorporating a company. Naïve, a startup that packages business infrastructure behind an API for AI agents, has crossed 30,000 developer customers within months of launch and just locked down a $28.5 million Series A led by Nexus Venture Partners, TechCrunch has exclusively learned.
CEO Sean Dorje says annual run-rate revenue has scaled 10x to the low double-digit millions over six months. The platform hands a prompt to tools like Cursor or Claude Code, and the agent handles the rest: forming a U.S. LLC, provisioning email inboxes, virtual cards, phone numbers, databases, and connecting to Stripe and QuickBooks. Humans still have to clear KYC hurdles and approve payments, but the orchestration layer governs budgets and requires sign-off for sensitive actions. Customers are running everything from faceless TikTok content channels—Dorje stumbled on one posting AI-generated boxing cats and dogs—to a fully autonomous rental car agency.
But the real money might be in making all those agents cheap enough to scale. The cost of calling expensive models, passing bloated context between tasks, and letting agents idle is brutal. Naïve is using part of the fresh capital to attack that directly. It’s building a model router that sends queries to the cheapest capable model while preserving and replaying already reasoned data, plus a memory system that surfaces business context on demand. More radical is a serverless runtime that runs agents in lightweight JavaScript environments instead of assigning each one a full virtual machine. Customers pay primarily when an agent is active, which makes deploying a swarm of them economically viable.
That serverless inference play is already the company’s fastest-growing demand driver. Dorje notes it’s drawing interest from enterprises, though he won’t name names. The trajectory is worth watching because it flips the value proposition. Developers may come for the one-click company formation. They’ll stay—and enterprises will pay—if Naïve can meaningfully slash the recurring compute bill for running agent fleets. The ten-person team plans to hire researchers and build out that infrastructure. Whether a startup this small can outmaneuver cloud giants on inference optimization is an open question. But 30,000 developers clearly don’t want to fill out incorporation paperwork ever again.
💡 Key Takeaways
- Naïve's serverless runtime ditches full VMs for lightweight JavaScript environments, so customers only pay when an agent is actually working—a direct attack on the biggest cost in autonomous business operations.
- The platform's growth is being driven in part by AI automation agencies reselling agents to small businesses, suggesting a B2B2B distribution model is emerging in the agent economy.
- Enterprises are showing interest in Naïve's inference optimization tech, which could transform the company from a startup incorporation tool into a cost-saving infrastructure layer for any business running agent fleets.
Keep reading: See related articles below for more coverage on this topic.
Get smarter about AI
The sharpest AI news, curated daily. Delivered free to your inbox.