Doug Schosser brings KeyBank AI oversight to Northwest Bancshares

TheCUBE's interview with Doug Schosser traces a path that is becoming common in regional banking: a finance leader trained inside a much larger institution — in this case KeyBank National Association — arrives at a smaller lender and brings not just software but a method for deciding what gets scaled. Schosser's point is that the size and scale of the two firms are very different, yet the discipline of screening AI proposals travels well. At Northwest Bancshares, that discipline takes the form of three tests applied to every AI proposal: customer impact, internal efficiency and risk reduction.

The Workiva platform Schosser carried over illustrates how the mechanics work in practice. Its years of document history serve as raw material for training agents on routine roll-forward tasks, which in his framing frees the human workforce for detailed explanations and connecting the dots. But he ties that payoff to a precondition: accurate, curated data, which he describes as a large orchestration problem spanning finance, information technology and the lines of business.

What the interview leaves open is whether the oversight model can match the pace of the tools. Schosser's own comparison — that AI receives less scrutiny than a junior analyst doing similar work — suggests the discipline he is importing is still catching up with what the software can already do. For regional bank finance chiefs weighing similar moves, the practical question is whether data orchestration across three functions can be assembled before the next wave of agent capabilities arrives, and whether vendors like Workiva keep governance at the center as they launch specialized AI agents.