Unified pipeline
Daily card and account transactions streamed from the core banking system
Banking · Qatar
Smart spend insights
Spending analytics embedded inside a leading retail bank's mobile and web banking — every transaction categorized, every pattern visible, refreshed through the day.
30% fewer statement-related calls to the contact center
+18% daily active users after launch
1.4× time-on-app
92% of transactions category-confirmed by early adopters
The challenge
What was in the way.
- Monthly e-statements listed raw transactions with no view of spending patterns.
- Cardholders phoned the contact center to ask where a charge came from — avoidable volume at scale.
- The bank's digital strategy called for personalized, value-adding services to differentiate from competitors.
What we built
The system, in brief.
Merchant intelligence
Every merchant mapped to its category; unknown merchants grouped by an ML model over names and locations.
Personal semantic layer
Spend aggregated by category
Embedded dashboards
Interactive charts inside the banking apps — category breakdowns
Outcomes
What changed.
- Statement-related contact-center calls fell 30% during the pilot — roughly 20,000 calls a year avoided.
- Daily active users rose 18% after the dashboard launch; time-on-app multiplied 1.4×.
- Early adopters confirmed categories on 92% of transactions, feeding the model's learning loop.
- The semantic layer opens a cross-sell runway for savings and round-up products.
Client referenced by sector and country · detailed references on request