AfterQuery hits $3.2B: expert work data is the product

AfterQuery is reportedly valued at $3.2 billion, five months after announcing a $30 million Series A at $300 million in April. Forbes first reported the round; TechCrunch says Y Combinator partner Gustaf Alströmer calls it the fastest any startup has gone from launch to unicorn in YC's history (TechCrunch). The founders are 22 and 23, out of the Winter 2025 batch. Round size was not disclosed.

What actually happened

AfterQuery trains models and agents on how professionals do their jobs — the company describes it as encoding the patterns, decisions and reasoning of the world's best practitioners. In practice that means paying doctors, lawyers and other specialists to produce training data, the same lane Scale and Mercor operate in. In April it said it had reached a $100 million annualized revenue run rate. Named customers include Nvidia, Legora and Motif Technologies, a Korean AI lab.

Why expert work data matters for your business

Strip the valuation away and the thesis is one sentence: the bottleneck on useful AI is not model quality, it is a recording of how competent people actually decide things. Labs are paying nine figures a year for that recording because they cannot generate it.

You already have a version of it, and you are almost certainly throwing it away. The quote your best estimator built and the three assumptions they changed before sending it. The support ticket that got escalated and the reason. The invoice someone rejected and what was wrong with it. That is the reasoning trace of your business, and in most small companies it exists as a decision made in a person's head, a number typed into a field, and nothing in between.

The practical move is not to sell data. It is to stop discarding the intermediate steps. When someone overrides a default, capture why in a structured field instead of a comment box nobody reads. When a workflow branches, log the branch and the input that caused it. Six months of that is the difference between an agent that guesses at your process and an agent you can actually evaluate against real decisions.

Be honest about the ceiling, too. AfterQuery is worth $3.2B because it aggregates expert judgment across an entire industry. Your dataset is small and specific — which makes it worthless to a lab and extremely valuable to exactly one company. That is the trade. Build for your own use, not for an exit.

Sources: TechCrunch, Forbes.