Data-Driven Approach to Process Automation in Banking
summary
Argues that automation readiness depends on planned data use and not just tooling, showing three bank processes and one delivered AI-based OCR replacing an out-of-date payment-order pipeline where the vendor hadn’t updated for a year. Notes modern cash counters yield 15–20 distinct data points per transaction, DMS PDFs often block extraction, and small feasible pilots beat two-year initiatives. Demonstrates operational gains via process mining + generative AI that can surface end-to-end bottlenecks and produce a full process analysis in about an hour.
AI-written summary · about
chapters (14)
- 0:00Data-driven banking automation
- 1:06Speaker background and Cycle Systems focus
- 2:58OCR payment processing failure
- 5:53AI in business misconceptions
- 7:17Automation beyond tools
- 8:44Banking machine automation evolution
- 11:22Digitalization vs transformation
- 13:10Automation project vision
- 15:45Top-down with bottom-up data
- 17:10Choosing automation projects
- 20:27Employee role in automation
- 21:51Data core of automation
- 24:01Process mining and analytics
- 26:11Q&A
related talks
31mBanking Transformation with AI: Extracting Value and Pioneering Change
37mApproaching AI in Banking
22mHow AI Implementation Transformed Financial Services in 2025
16mMy Bank Meets AI: Smarter Decisions Beyond Fancier Spreadsheets
09mAutomating cognitive tasks to increase business value
38mFrom Paperwork to Pipeline: Autonomous Mortgage AI
36mHow can AI turn data overload into smarter decision-making?
29mAI - Digital (r)evolution in banking