Does expertise influence data validation work?  We explore this question by observing financial data validation work by a group of data scientists (DS) vs. a group of accounting (AC) graduate students.  We found interesting differences between the DS and AC groups, in terms of their use of one- versus two-dimensional data visualizations, use of multiple visualizations, use of business versus data rules, navigational patterns, and others.  These findings suggest that domain expertise plays a critical role in data validation for financial AI development.