← All work01 · BANKING · OPS
Performance culture and AI tooling delivered up to 40% productivity uplift in a 1,200-person operations department
Context
Post-pandemic cost-to-income pressure had stalled a long-running efficiency program. The bank had invested in digital tooling and early GenAI experiments but couldn't point to measurable throughput gains. Adoption was uneven, performance management inconsistent across cluster leads, and the productivity ceiling looked stuck despite the investment.
Problem
Leadership needed a measurable uplift inside a fixed program horizon — and the operating muscle to keep delivering after the consultants left. Three things were missing: a shared performance baseline, a coaching cadence at team-lead level, and a focused AI rollout with real adoption mechanics behind it.
Approach
- 01
Diagnostic
Shadowed teams across the department, ran ~40 interviews, segmented operations by process type and uplift potential. Identified eight high-leverage process clusters and ranked them by speed-to-impact.
- 02
Performance culture
Installed a shared management system across all clusters — common KPIs, daily 15-minute huddles, transparent dashboards visible to every team member. Trained ~60 team leads on coaching rather than supervision. The unlock wasn't the metrics; it was making team leads owners of their numbers.
- 03
AI-enabled productivity layer
Selected and rolled out a focused stack: GenAI for document drafting and summarization, intelligent document processing for inbound mail, knowledge retrieval for the service desk. Piloted in three clusters with pre/post baselines and a control group. Built an ambassador network of ~25 power users to drive peer-to-peer adoption — the only mechanism that consistently sticks.
40%
peak productivity uplift in pilot clusters
8
high-leverage process clusters identified
Outcome
- Up to 40% productivity uplift in the best-performing clusters
- 15–25% average uplift across the broader program
- Daily huddles and dashboards institutionalized across all 1,200 FTE
- Pilot clusters became the bank's internal proof points for the next wave of AI investment
- Three of the eight clusters reached their 12-month productivity target inside six months
Stack
Power BIMS TeamsExcelMiroConfluenceMicrosoft CopilotChatGPT