Search⌘ K
AI Features

The Case Study

Explore the loan document flagging case study focusing on automating underwriting checks in a regulated bank. Learn to interpret customer briefs, handle regulatory constraints, and design systems that support underwriters by identifying document inconsistencies efficiently and reliably.

Case studies in an FDE interview rarely arrive as tidy specifications. An interviewer describes a customer, a problem, and a few constraints, then watches how we work through it. This case follows that shape. It places us as a Forward Deployed Engineer embedded with a regional bank that wants Claude to help its loan team.

This lesson presents the case as a customer brief. It describes the bank, how the work is done today, what the bank is asking for, the rules it operates under, and what the engagement covers. Working through the case, from discovery to design to the hard conversations, comes after the brief is clear.

The case begins with the team that does the work.

The bank and its underwriting team

Northfield Regional Bank is a mid-sized bank that serves customers across three states. Alongside checking accounts and mortgages, it offers personal loans, which are unsecured loans of a few thousand to tens of thousands of dollars that customers use for things such as consolidating debt or paying for a home repair.

Every personal loan application goes through underwriting, the process of reviewing an applicant’s documents and financial history to decide whether to lend, and on what terms. The people who do this work are underwriters. Northfield’s personal loan team has nine underwriters and handles about 2,400 applications a month. The bank promises applicants a decision within two business days, ...