Back to portfolioGet in touch

Case Study 04 · Bringing conversational AI into an enterprise banking product

When AI isn’t a feature — it’s part of the product.

Conversational UX patterns for consumer banking, designed before “ChatGPT for banking” was mainstream — part of NCR's $100M+ enterprise financial-services engagement.

Enterprise
Banking
Conversational
AI
Multi-brand
North Bank + First Digital
$100M+
Broader enterprise engagement
01 · The problem

Banking customers shouldn’t have to understand the bank’s architecture.

“Show me my transactions.”

“Help me with my card.”

“Pay someone.”

“What’s happening with my account?”

The opportunity was to make those intents accessible through a conversational interface — without separating the AI from the underlying banking experience.

02 · The product vision

Make conversation a first-class banking interaction.

Not
Banking app + chatbot
But
Banking app with intelligence embedded into the workflow
03 · Experience model

Intent to action, in one conversational loop.

Conversation
Intent recognition
Banking capability
Action
Confirmation
Next best action
Example
“Show my transactions.”
Transaction lookup
Results
“Do you want to view a specific transaction?”
04 · The interface

Ask Kai — conversational banking as a first-class surface.

01 — Conversational entry point, integrated directly into the banking app.
01 — Conversational entry point, integrated directly into the banking app.
02 — Intent-driven transaction shortcuts (Show Transactions, Cards, Pay someone, Help).
02 — Intent-driven transaction shortcuts (Show Transactions, Cards, Pay someone, Help).
03 — Multi-brand conversational UI running on the shared NCR Design System.
03 — Multi-brand conversational UI running on the shared NCR Design System.
05 · Multi-brand product system

One conversational model, deployed across brands.

Core conversational model
NCR design system
North Bank — different branding / customer context
First Digital — different branding / customer context
06 · Enterprise constraints

Product decisions had to balance seven things at once.

Customer trustSecurityFinancial accuracyBrand consistencyAccessibilityTechnical feasibilityEnterprise scalabilityHuman assistance
07 · Why it matters today
The technology changed. The product problem didn’t.

AI products succeed when intelligence is integrated into the customer’s actual workflow — not when a chatbot is simply added to the interface. The same principle applies to industrial, operational, and enterprise AI: the value comes from helping people make better decisions and complete meaningful work.