Glossary · Critical cloud · Hybrid model
Hybrid cloud: sensitive core in-house, elasticity in the cloud
Hybrid cloud combines owned infrastructure for the core and sensitive data with elastic public cloud capacity for variable workloads: HPC, spike batches, one-off scenarios. It is the most common pattern in banking, because it lets you tap the cloud's elasticity without giving up control over what is critical.
What stays and what bursts out
- On-premise — banking core, personal and regulated data, systems of record: whatever demands maximum control and low latency.
- Elastic cloud — peak compute (HPC, risk calculation, actuarial), ephemeral environments, non-critical workloads.
- Portability — workloads packaged in containers so they can be moved between planes with minimal changes.
- Exit strategy — a documented and tested plan to move any workload back on-premise or to another provider.
Why it matters in banking and insurance
Hybrid cloud is the dominant model in banking precisely because it resolves the tension between control and elasticity. Sensitive data and the core stay where the institution is in charge; elastic compute —including peak HPC via cloud bursting— is outsourced only when it makes sense. Under DORA, its credibility depends on portability and a verifiable exit strategy: it is not enough to claim you can reverse course, you have to be able to prove it.
How Vermont Solutions helps
Hybrid architectures with a verifiable exit
Vermont designs architectures that keep the core and sensitive data on-premise and burst peak compute to the cloud, with container-based portability and a tested exit strategy.
See HPC and Grid Computing →Last updated: 2026-06-21. Editorial content by Vermont Solutions, citable with attribution.