I Founded a 20-Person AI Company With Big Competitors — Here’s How to Out-Ship a Team 3 Times Your Size

You know the competitor with three times your people and three times your budget. Sooner or later, that gap makes you wonder, “Do I have to hire the way they hire to keep up?”

Most of the time, you don’t.

Y Combinator CEO Garry Tan told CNBC that companies are reaching as much as $10 million in revenue with teams of fewer than ten people. “You don’t need a team of 50 or 100 engineers,” he said. Base44, a bootstrapped app builder with eight employees, sold to Wix for $80 million six months after launch.

Small teams can win with an AI-first playbook

As the founder of a 20-person AI company with much bigger competitors, I believe AI can help small teams punch above their weight. But this doesn’t mean using AI for its own sake.

Instead, the best AI-native playbook involves validating AI performance data, holding AI accountable and using AI in ways that don’t bury humans in time-consuming new review tasks.

The strategies below require discipline rather than large teams or big budgets, and you can start them today.

Verify AI-reported performance data

Every AI tool gives you a dashboard of hours saved and output produced. I ignore those numbers because they measure volume. A rising volume line can mean progress or a growing pile of busywork.

When the code analytics firm GitClear studied 211 million changed lines of code, it found copy-pasted code exceeding refactored code for the first time in 2024 and tied the shift to AI coding assistants. More lines shipped, and more of them were duplicates someone will have to maintain.

Tracking how quickly you hit roadmap goals is a better way to assess AI’s impact. A feature slated for late in the third quarter that ships in mid-April has pulled you roughly 170 days forward. Log planned versus actual dates for your next three roadmap items, and keep ignoring the dashboard.

Build prototypes instead of lengthy feature documents

In my experience, the fastest agent-era teams spend more time deciding and less time documenting.

At my company, we used AI to eliminate lengthy proposal documents. When someone has a new idea, they use AI to prototype it fast. Then the team plays with it and makes a go/no-go decision in an hour. You can do this, too.

Write a one-page contract before an agent writes code

Once a feature is approved, we give the coding agent a lean, one-page document naming what the feature does, the data it touches, the interfaces it exposes and the ways it can fail.

A specific, well-written contract dramatically increases code quality and reduces your time-to-ship. A poorly written or missing contract slows you down: Auditing and fixing bad code costs more than the agent saved.

A 2025 Faros AI study of 10,000-plus developers found teams with high AI adoption completing 21% more tasks while code review time rose 91% and bug rates rose 9%.

Require the one-page contract this week for any agent-generated work.

Let automated gates and agents do preliminary reviews

The top frustration with AI tools in the 2025 Stack Overflow Developer Survey, named by 66 percent of developers, was solutions that are “almost right, but not quite.”

Most teams counter these buggy solutions by adding more human review at every stage. I think that instinct can be counterproductive.

Consistent automated testing, built as a series of deterministic gates, frees humans from the first line of review. That can mean requiring a compiler, type checker, linter and a full test suite in front of every merge, then letting agents attack each change like an adversary.

We do this at my company, and the result is cleaner code. My development team spends more time architecting tests and defining outcomes than combing through code. Make one gate non-negotiable, starting with your next merge.

Rank work by what it changed for users

AI made output cheap, and cheap output breaks scoreboards. Commits, tickets, drafts and decks inflate overnight while nothing changes for your customers.

I call this comfort work. It looks and feels productive but doesn’t advance any goal, and AI multiplies it because iterations are nearly free.

Attach a verifiable test to every goal that a stranger could check. “Checkout conversion crossed 4% and held for 30 days” is verifiable. Anyone can look it up. “Improved the funnel” is vague and can’t be validated.

Rank each project against these tests instead of by number of artifacts, and your team learns fast that shipping only counts when it affects a goal.

Use AI for your hardest decisions

The widest gap between a small team and a big one is how quickly they can make good decisions. Big teams have more data and expertise, so they more often choose the right path. Small teams can close the gap by using AI for decision support.

To get an unbiased answer and catch AI fabrications, ask several models the same question. Each has its own map of the world, so the places where they agree suggest a consensus.

Then have one model, or a skeptical teammate, try to discredit that consensus. Consensus tells you what’s popular. Verification tells you what’s true.

At my company, questions that once took a senior researcher a week now come back that morning. Run your next category bet or pricing call through this before the meeting that decides it. It is the throughline of what we’re building: as more buyers and their AI assistants ask whether they can trust a recommendation, the advantage goes to whoever can verify what the model only guesses at.

Discipline is the moat

A competitor with three times your people carries roughly nine times the coordination cost. Fred Brooks named the law a half-century ago in The Mythical Man-Month. “Adding manpower to a late software project makes it later.”

Every new hire coordinates with everyone already there, so working relationships multiply faster than headcount. Five people have 10 of them. A hundred have nearly 5,000.

Shopify runs hiring on this math. CEO Tobi Lütke told his teams to demonstrate why AI cannot do a job before asking for more headcount.

The teams that rebuild around these six shifts win the decade. The ones that wait will spend it catching up to teams a third their size.