Community banks are losing good commercial deals to speed. A borrower package lands as 40 PDFs, spreading eats three days, and by the time the memo reaches committee a credit union or non-bank lender has already sent a term sheet.

This lab walks the commercial credit workflow end to end and marks the seven steps where AI is genuinely ready today— intake, spreading, cross-document conflict detection, covenant extraction, and memo drafting with source traceability— and the three where it is not, including final risk rating and adverse action reasoning.

Led by a former commercial credit officer who underwrote 3,000+ C&I, CRE, and ABL deals at Santander and CoBank, this is a practitioner's view, not a technology overview. Every claim is tied to a real file example. Attendees leave with a one-page workflow map and a review protocol for checking AI-drafted spreads and credit memos before committee. We close by pulling apart a redacted AI-generated memo together and finding what it got wrong.

Learning Objectives:

• Identify the seven steps in the commercial credit workflow where AI is production-ready today and the three steps where it is not.
• Evaluate AI-generated spreads and memos for the failure modes that matter in credit: unsourced figures, numbers left stale after a document is replaced, misconsolidated related entities, and conflicts across the tax returns and aging reports.
• Assess which parts of your own borrower document intake create the most rework and what has to be fixed upstream before any automation produces output your committee will trust.