Banking ops teams sit on enormous amounts of customer data - transaction histories, card states, program details, spend patterns but accessing any of it requires either a technical resource or a pre-built report. Neither is fast enough when a customer is on hold or a credit decision needs context. Engage AI was built to close that gap. Connected directly to VegaPay's CCMS data layer, it lets support agents, operations teams, and credit analysts ask questions in plain text - "How many customers are in the Aura credit card program?" and get structured answers instantly, with the option to inspect the underlying SQL or switch to a graph view. The design challenge was significant: the users aren't data analysts. They needed an interface that felt like a conversation, not a query tool but with enough transparency for teams that needed to validate results before acting on them. I designed Engage AI end-to-end from the brand identity and hero concept to the query interface, result states (value, table, graph), and SQL visibility toggle. The product was built internally at VegaPay and demoed to clients alongside CCMS as part of the broader platform story.