- AfterQuery’s valuation has reportedly jumped from $300M to $3.2B in five months
- Founders Spencer Mateega and Carlos Georgescu joined YC just 18 months ago
- AfterQuery’s revenue reportedly grew from $100M to hundreds of millions since April
AfterQuery has reportedly raised a Series B at a $3.2 billion valuation, according to Forbes, one of the fastest valuation jumps the AI startup market has seen this year.
The San Francisco-based startup was valued at $300 million when it announced a $30 million Series A in April. Five months later, the new figure is more than a tenfold increase and, according to Y Combinator partner Gustaf Alströmer, the fastest run from launch to unicorn status in the accelerator’s history.
AfterQuery is reportedly profitable and has already lined up a lead investor for the round, though the company declined to comment.
From YC cohort to $3.2 billion startup
Spencer Mateega and Carlos Georgescu, high school friends who founded AfterQuery in 2025 and joined Y Combinator’s Winter 2025 batch, were 23 and 22, respectively, when the new valuation was reported.
The company did not set out to build a data business. Mateega and Georgescu first tried building AI agents for financial workflows, then found that existing models kept failing on complex, nuanced professional tasks — not because the models couldn’t reason, but because they had never been trained on how professionals actually work. That insight led to the pivot.
AfterQuery now pays experts across software engineering, finance, law, and medicine to capture how they reason through problems, then turns that into datasets and reinforcement-learning environments AI labs can train on.
Why AI labs are paying for human expertise
Early AI systems learned plenty from public web data. That well is running dry: as frontier models exhaust what’s freely available online, labs are turning to costlier, domain-specific data that shows how skilled professionals actually solve hard problems.
AfterQuery’s customers reportedly include Nvidia, legal AI company Legora, and Korean AI lab Motif Technologies, and its data has been used in Nvidia’s Nemotron models. The company has also worked with Thinking Machines Lab, the startup founded by former OpenAI chief technology officer Mira Murati.
The company said it had crossed $100 million in annualised revenue when it raised its Series A in April. Mateega posted on X in July that the figure had grown to “hundreds of millions” — and said the company runs its own internal post-training pipeline, training models on its data first so labs can see proof of quality before they evaluate it themselves.
A crowded AI data market
AfterQuery isn’t the first to show investors the value of human expertise here.
Mercor, which Tech Funding News reported raised $350 million at a $10 billion valuation in October 2025, has since gone on to seek a valuation near $20 billion. Deccan AI, which supplies post-training data and evaluation work through a network of experts, raised $25 million in a Series A in March.
For AfterQuery, the size of this round suggests investors think expert reasoning data is becoming more than a supporting layer for AI development. As models push deeper into professional work, teaching them how experts actually think may turn out to be one of the more valuable, and most expensive, pieces of the AI stack.
