Casca AI: Building an AI Native Loan Origination Platform

By Khatidja Vaiya, Jim Ellis, Ilana Stern
2026 | Case No. E932 | Length 11 pgs.
This case follows Stanford GSB graduate Lukas Haffer and cofounder Isaiah Williams as they build Casca AI, an AI-native loan origination platform that tackles the complexity of U.S. Small Business Administration (SBA) lending. It traces Casca’s journey from problem validation at Y Combinator, through a derailed “platform” partnership in Europe, to a narrow but high-impact wedge in SBA lending with community bank Havenbank that dramatically improved application conversion and processing times for small business borrowers. As Casca “ladders up” to win Birchstone Bank, the country’s largest SBA lender, Haffer confronts intertwined strategic decisions: whether to deepen into a system-of-record SBA platform or broaden horizontally across loan types, and whether to accept a preemptive Series A from bank-backed Canapi Ventures and customer-investors or pursue top-tier Silicon Valley VCs.

Learning Objective

The case invites students to evaluate how early-stage fintech founders should choose vertical depth versus horizontal expansion, and how to weigh strategically valuable but potentially conflicted customer capital against more traditional venture funding.
This material is available for download by current Stanford GSB students, faculty, and staff only. For inquires, contact the Case Writing Office. Download