Brazil's Courts Are the World's Hardest Training Data. That's Why Enter Is LatAm's First AI Unicorn.

Brazil processes roughly 84 million active lawsuits at any given point — more than any other country on earth. Most founders look at that number and see a dysfunctional legal system. Founders Fund looked at it and wrote a check.

Enter, the São Paulo legal AI company founded by Mateus Costa-Ribeiro in 2023, raised $100 million at a $1.2 billion valuation in May 2026, tripling in eight months and becoming Latin America's first AI unicorn. The round was led by Founders Fund, with Sequoia Capital, Ribbit Capital, ONEVC, Atlantico, and Kaszek participating. Between the September 2025 Series A — itself co-led by Founders Fund and Sequoia at a $350 million valuation — and this round, revenue increased more than 10x and the client base tripled. Clients now include Itaú, Santander, Mercado Livre, Nubank, and Airbnb: companies processing high-volume litigation not because they choose to, but because Brazilian consumer and labor law structurally guarantees it.

The moat is structural rather than technical. Enter's product — EnterOS — conducts the judicial process from beginning to end: case intake, procedural tracking, response drafting, settlement calculation, outcome prediction. That capability alone doesn't make the company defensible. What does is the training data underlying those predictions. Brazilian jurisprudence carries decades of layered precedent, a labor law framework that generates millions of cases annually, and a consumer protection culture that routes virtually every dispute to court rather than arbitration. Training on 300,000 new cases per year in that environment produces legal AI that cannot be replicated by a US or European competitor — not because of model architecture, but because the training environment doesn't exist anywhere else. The complexity of Brazil's legal system, which most foreign investors treat as a reason to stay away, is precisely what makes Enter's position defensible.

Mateus Costa-Ribeiro's background is unusually relevant to the product. He became Brazil's youngest practicing lawyer at 18, passed the New York Bar at 20, and graduated from Harvard Law before leaving a Stanford MBA to found Enter. That trajectory is worth noting less for the credential stack than for what it means about founder-market fit: he built automation for a system he practiced in, with an intuition about what breaks down at volume that outside observers cannot develop through desk research. Founders Fund and Sequoia, both notably selective in their international bets, co-invested at two consecutive rounds. That consistency of conviction is itself a signal.

The question that will shape Enter's next phase is geographic. Brazilian legal AI is highly specific — the CPF framework, TST precedents, labor law traditions, and consumer code enforcement mechanisms are not portable. Mexico runs a different system. Colombia another. Argentina a third. The data moat that makes Enter defensible in Brazil also constrains simple geographic replication. The company can build regional presence through local data accumulation and system configuration, but it cannot deploy the same model cross-border. How Enter manages that tension — whether the $1.2 billion is the foundation of a continental legal AI platform or a highly defensible single-market business — will be determined by how aggressively it moves in the next eighteen months before any local competitor develops comparable data depth.

Enter becoming LatAm's first AI unicorn is less a story about AI and more about what happens when the hardest operating environment becomes the best training set. The insight isn't "Brazil has a lot of lawsuits." The insight is that difficulty at scale is the input for moats at scale — and investors who understood that early are now marked up 3x in eight months.