Polsia: Solo Founder Hits $1M ARR With AI-Run Companies

How Polsia Turns AI Agents Into Autonomous Company Operators

Ben Broca is the solo founder of Polsia, an AI platform that builds and runs companies autonomously. In a live conversation with Swyx on Latent Space, Broca demonstrated a system managing over 1,000 companies simultaneously — handling engineering, marketing, cold outreach, social media, and meta ads — all without human employees. The platform crossed $1M ARR on the day of the interview, roughly one month after launch.

The core product loop: Users give Polsia a business idea, and an AI CEO agent wakes up every night to evaluate the state of the company, decide what to work on, execute tasks, and send the founder a morning email summarizing what happened and what comes next. “Every night it wakes up and there’s a CEO sort of like instance that will decide based on if there are any bugs, how’s the business doing, do we have any paying customers, and it will decide what to do and then execute on it.” Users guide the AI via email replies or a dashboard chat, creating a daily feedback loop.

On model selection and intelligence: Broca runs Claude Opus 4.6 as the primary reasoning model for Polsia’s CEO agent, despite the cost. “This is the agent that’s going to decide on strategy, that’s going to advise the user on what is the best plan to get to a successful company. So I think it’s important to give it the best reasoning.” The platform demonstrated surprising emergent behavior during the demo — identifying a test account by recognizing the founder’s own email pattern and reasoning about the relationship.

On stripping complexity for adoption: Rather than asking users to connect their GitHub, email, or ad accounts (which creates friction and skepticism), Polsia provisions everything: email addresses, Render web servers, Neon databases, Stripe accounts, and GitHub repos. “Let me strip out all this complexity, let me provision everything for them. The hard part was deciding what not to build.” Broca compared this to an Apple-like approach versus the Android-style openness of tools like OpenClaw.

On the business model: The $50/month subscription roughly breaks even on AI costs. Real revenue comes from a 20% cut on business revenue collected through Polsia’s Stripe accounts and a 20% cut on ad spend managed by the platform. This aligns incentives: Polsia makes money only when the user’s business makes money.

On using AI to run Polsia itself: Broca practices what he preaches — AI agents handle customer support, respond to investor inbound, find and fix bugs, and even build features based on user requests. “For any person I need to hire, I can actually build agents that are going to do the work. If I’m going to sell the promise of an AI that builds and runs companies, if I’m myself using the service to run Polsia, I’m sort of living proof.”

5 Takeaways From Polsia on AI-Powered Entrepreneurship

• Autonomous AI CEO as daily operator - Polsia’s nightly execution loop (evaluate, decide, execute, report) turns the AI from a chatbot into an operational partner that drives business forward without human intervention
• Provision everything, ask nothing - Removing onboarding friction by provisioning infrastructure instead of asking users to connect existing accounts was the key product decision that unlocked growth
• Revenue alignment over token reselling - Breaking even on subscriptions and taking 20% of business revenue creates an incentive structure where Polsia succeeds only when users succeed
• 15 messages per day per user - Users treat the AI CEO as a real co-founder, averaging 15 daily messages to discuss strategy, features, and direction for their autonomous businesses
• Solo founder at scale - One person managing 1,100+ autonomous companies and crossing $1M ARR demonstrates that AI agents can replace entire teams, not just individual tasks

What Self-Running Companies Mean for AI-Powered Organizations

Polsia represents a new category: platforms that don’t just assist with business tasks but autonomously operate entire companies. The most striking insight is the engagement — 91,000+ human messages across the platform suggests users aren’t just delegating and forgetting. They’re actively co-founding with AI, guiding strategy while the system handles execution. If this model scales, the question for every business isn’t whether to adopt AI agents, but whether a solo founder with the right AI platform can outcompete entire teams.