Designing and operating a fully automated public cloud platform for digital banking
Key takeaways
- A public cloud platform was designed from the ground up to support a regulated banking environment.
- Infrastructure provisioning and deployments were fully automated, reducing operational risk and setup time.
- The platform now supports a large-scale banking system built on hundreds of virtual machines and containerised services.
Vodeno, an international banking start-up, required a cloud-native foundation capable of supporting a complete digital banking ecosystem. The platform needed to handle transactional processing, API exposure, integration services, monitoring, and regulatory constraints, all while remaining flexible enough to evolve with product expansion.
The objective was not limited to infrastructure migration. The organisation required a scalable, automated, and operationally mature platform built entirely in the public cloud. It also required a robust communication backbone (data bus) and a structured tool ecosystem supporting both Linux- and Windows-based workloads.
This meant combining architectural design, infrastructure engineering, DevOps processes, and ongoing platform operations into a single coherent delivery model.
The challenge
Designing a cloud platform for banking involves a unique combination of constraints.
- The system had to comply with regulatory expectations typical for financial institutions.
- Cloud selection required objective evaluation of major public providers.
- The architecture needed to support high transaction volumes and service-based communication.
- Infrastructure provisioning and releases had to be automated to avoid manual operational bottlenecks.
- Monitoring and business process visibility had to be embedded into the platform from the outset.
The complexity was amplified by the need to support both platform construction and continuous evolution in parallel with product development.
Solution overview
The engagement began with functional testing and evaluation of the three largest public cloud providers. Based on defined criteria, including scalability, service maturity, cost structure, and operational alignment – Google Cloud Platform was selected as the target environment.
A complete platform architecture was then designed to meet business and regulatory requirements. The architecture combined infrastructure-as-code practices, container orchestration, and service-based communication patterns.
The solution incorporated:
- Terraform for infrastructure provisioning,
- Kubernetes and Docker for containerised workloads,
- Packer and Ansible for configuration management,
- GitLab CI and Jenkins for automated pipelines,
- ELK stack (Kibana, Elasticsearch, Logstash) for observability,
- Apache Kafka as the data bus layer,
- Kong EE and NGINX/F5/Squid for API management and traffic control,
- Cassandra and PostgreSQL for data persistence,
- Java/Spring, Node.js, and Python for application services.
Automation was central to the design. Infrastructure, environments, and deployments could be created and replicated in a controlled, predictable manner.
How the work was executed
The project combined platform engineering with application-layer collaboration. Cloud architects and DevOps engineers worked alongside application architects and developers to ensure that the infrastructure supported actual business use cases.
Infrastructure was provisioned entirely as code, enabling repeatability and consistency across environments. Kubernetes clusters hosted containerised services, while the communication backbone based on Kafka ensured reliable service-to-service interaction.
Continuous integration and deployment pipelines automated the release process, reducing dependency on manual execution. Monitoring mechanisms were implemented to provide visibility into both infrastructure health and business process execution.
Beyond initial construction, the team remained responsible for ongoing maintenance, optimisation, and incremental modifications of the Google Cloud environment. The platform operates as a living system, evolving alongside product development.
Operational results
The outcome was a fully automated public cloud platform supporting a complete banking solution.
Key results included:
- automated provisioning of cloud environments,
- deployment of several hundred virtual machines and containerised services,
- stable support for a full digital banking system,
- integrated monitoring and resource control mechanisms,
- a scalable communication backbone enabling service growth.
The platform provides a structured operational model rather than an ad hoc infrastructure setup.
Business value
By designing and operating a cloud-based banking platform from first principles, Vodeno gained a technical foundation aligned with its long-term growth strategy. Automated infrastructure reduces setup time for new environments and limits configuration drift.
The combination of container orchestration, infrastructure-as-code, and integrated monitoring improves operational predictability and resilience. Regulatory readiness is supported by structured logging, traceability, and controlled deployment processes.
The platform allows product teams to focus on developing banking services while infrastructure remains stable, scalable, and continuously maintained.
By building and maintaining a fully automated cloud platform, Vodeno established a scalable and operationally mature foundation capable of supporting a complete digital banking ecosystem.