Inside the Top 5 Advisory Firms: AI Builds the Model. Who Checks It?
Inside large advisory firms, AI can build leveraged buyout (LBO) and discounted cash flow (DCF) models faster than teams can review them, moving the bottleneck from building models to trusting them.
Drawing on Crunched's work with eight of the ten leading advisory firms, PwC among them, Michael walks through the three stages of adoption these firms go through: error checking first, then generation capped at what one reviewer can check in a sitting, then running groups of AI agents with AI-assisted review at scale. He shows where the most advanced teams are stuck today.
Attendees leave with the review practices that separate teams getting more analysis from AI from teams getting more unchecked models, and the risks that follow when that line is crossed.
Other Sessions
A hands-on day building reusable finance tools with Excel, VBA, Power Query, Power Automate and Copilot to automate consolidations, reconciliations and burst reporting every month.
Sloka shows how a three-statement model feeds a valuation, what models can and can't tell you about value, and where valuations most often go wrong, using anonymised examples.
Session title and details coming soon.
Danielle shows how nested arrays solve real modelling problems, and how review habits must change when one cell can hold a whole depreciation profile or scenario path.








.png)














.png)
.png)





.png)


.png)















