Everyone is being told that AI means smaller teams. Omer Kaplan started his new company believing exactly that, planned to stay at ten people, and then changed his mind.
Kaplan co-founded ironSource, ran its revenue for more than a decade, and watched the company list on the New York Stock Exchange in June 2021 at an $11 billion valuation. Unity bought the business the following year, and he left in January 2024. He is now co-founder and CEO of ZyG, which came out of stealth in March 2026 and has raised $118 million from Bessemer, Viola, Lightspeed and Accel.
He does not argue with the math everyone else is doing. He agrees that one experienced engineer with an army of agents does what ten people used to do. What he disputes is the conclusion. You can take that saving home or you can spend it, and he decided to spend it, because what he is building is thirty startups running at once.
The reason is in the other half of the conversation. For three years the winning move in e-commerce software was to pick one expensive problem and solve it well. Landing page generation. Creative production. Email marketing. Those founders are calling him now, after failing to raise a follow-on round, and his verdict is blunt. Either the customer builds it themselves with Claude, or somebody else does. There is no moat around one problem.
What he thinks there is a moat around is delivering an outcome. A good consumer brand today has two options and neither is real. Amazon will rent it demand without letting it own a customer. Shopify will sell it plumbing. Neither one takes responsibility for whether the brand actually grows. Kaplan wants to be the third thing, and he has run the play before: at ironSource, Supersonic took mobile games from developers anywhere, scored each one on cost per install and day-one retention, and pushed the ones that cleared the bar. Some of those developers were working an oil rig in Turkey. They became multimillionaires.
In this conversation with Mickey Haslavsky, Omer explains the screening machine that decides which consumer brands get scaled and why none of it runs on instinct, what a moat looks like when it turns out to be a sachet of gummies rather than a patent, why he broke with the revenue-share model he helped build in mobile games so the partner keeps the intellectual property, what happens to junior engineers in a market bidding only for senior ones, and why he has no fear at all of the second album.