PitchBook has used AI in data operations for years. Our hybrid AI/human intelligence model — AI-powered ingestion, classification, and entity resolution paired with human validation and primary research — allows us to source and structure data from hundreds of thousands of sources daily, many of which are not publicly available. In the AI era, the human research layer can become even more valuable because it provides the training signal and quality benchmark that help keep AI-driven outputs accurate.
Our data operations team realizes substantive efficiency gains each year. That said, AI's standalone margin contribution is difficult to isolate, while savings from the efficiencies we’ve realized have largely been redeployed into expanding coverage and new datasets rather than flowing straight through to margin. This is consistent with our philosophy of building our data moat for long-term growth.