The pollsters missed again; the market saw it first—and a Brazilian was at the center of it.

In the final stretch of the race for the White House, almost the entire polling establishment projected a Kamala Harris lead. They were wrong. Again.

On the prediction markets, though, Trump had already flipped. Numbers backed by real money, not opinions.

Luana Lopes Lara—a Brazilian mathematician and computer scientist trained at MIT, co-founder of Kalshi—spotted the shift earlier than almost anyone. “People are putting money behind what they believe is the most likely outcome,” she said. That single detail defeats any poll vulnerable to bias, bad sampling, or respondents who lie.

Luana didn’t come from the cradle of Wall Street. She grew up in Niterói, studied with obsession, competed in math Olympiads, danced at the Bolshoi, got into MIT, Harvard, Yale, and Stanford. She worked at Bridgewater, Five Rings, and Citadel before founding, at 22, a CFTC-regulated exchange. Regulated—something the election-betting world always avoided.

Kalshi raised $100 million in venture capital, including Sequoia, Schwab, and Henry Kravis. And it won, twice, its legal battle with the CFTC to operate election contracts in the U.S. In election week it became the most downloaded app on the App Store. The site passed 100 million visits. Total volume crossed $1 billion.

While polls crashed narratives, Kalshi’s curve adjusted probabilities in real time—like any liquid market that doesn’t tolerate error for long.

Bettors range from $100 hobbyists to positions in the tens of millions. In 2024 the company partnered with Susquehanna to secure liquidity. The logic is simple: liquid markets are hard to manipulate. “There are correction forces,” Luana says. And there’s solid literature showing that mispriced assets don’t survive.

Kalshi operates only for U.S. residents. For the rest of the world, the vacuum is filled by Polymarket, backed by Vitalik Buterin and the Founders Fund, running on USDT. It’s the parallel version—less regulated, but functional and global.

And the crucial point: Luana’s next step is obvious. “I want to launch markets for Brazil’s elections.” When that happens, Brazilian voters will have, for the first time, a realistic thermometer of expectations—something traditional polls failed to deliver even in the U.S.

The story isn’t just about Trump. It’s about a systemic error. And about a Brazilian who built the instrument that spotted it before anyone else.