Banks have it harder than most sectors because of Basel III and AML requirements. We're a regional bank doing transaction reconciliation and loan processing with 22% on-premise, 78% cloud split. The trick was identifying which processes actually touch regulated data versus metadata and reporting. Most reconciliation doesn't need the raw customer data, just flagged exceptions. That insight moved 60% of our workload to cloud where scaling costs dropped from $45K monthly to $18K. Regulatory reporting still runs on locked-down on-premise infrastructure with air-gapped networks. Total ROI hit positive at 9 months instead of the projected 16 months.

Healthcare here with similar constraints under HIPAA. We ended up with unattended bots running on-premise for patient records and PHI processing, while customer service automation and appointment scheduling runs cloud-native. The key was deploying UiPath's orchestrator in hybrid mode with policy-as-code governance. Compliance tracking happens real-time across both environments through centralized dashboards. Implementation took four months and required CISO signoff on every data flow, but now we're processing 80,000 claims weekly with zero breaches. Your KYC use case would probably need OCR and document extraction on-premise with cloud handling the workflow orchestration.

Our bank processes 150,000 KYC documents monthly and we're stuck between cloud-based RPA scalability and on-premise compliance requirements. Current setup is 58% on-premise because regulators want data residency guarantees, but cloud solutions like Automation Anywhere's SaaS platform would cut our infrastructure costs by 35-40%. The hybrid deployment model everyone talks about sounds perfect on paper, but actual implementation with regulatory reporting and audit trails gets messy fast. Enterprises running hybrid RPA in regulated industries - what's your architecture look like? How do you handle the sensitive data that can't touch cloud servers while still getting the elasticity benefits?