Most finance teams have already tried AI on something. The harder question is where it holds up once real numbers are involved, like drafting variance commentary or flagging reconciliation breaks before close. This session-length look at real use cases across ops and finance also gets into where a person still has to own the output. Yudhajit spent years on finance transformation work at Deloitte, including intercompany reconciliation for controllership teams, and now trains analysts in financial modeling and Python for finance. That mix of process work and hands-on modeling is a useful lens for telling apart the tasks worth handing to AI from the ones that only look automatable.