Case study – Energy
A hydropower operator's app, running on sovereign Swiss cloud.
Hydro Exploitation runs hydroelectric plants across the Valais. Its predictive-maintenance product, GuardHEX, flags anomalies in generator windings, turbines, transformers, and bearings – but needed a production home its own team could operate. 56k.Cloud containerised the application, built the DevOps pipeline, and deployed it to Exoscale on managed Kubernetes – in about ten working days.
The problem
GuardHEX flags anomalies in plant equipment – generator windings, turbines, transformers, bearings – before they become failures. But the tool its analysts used to track those anomalies through to resolution was a shared Excel file, and the application itself had no production home its own engineers could run. Hydro Exploitation wanted to modernise: move to public cloud, put the app on a proper DevOps footing, and have its own technical team able to operate and evolve it.
- Swiss energy operator – workloads and data to stay under European jurisdiction
- the customer's team must be able to own and run it, not just receive a black box
- reuse the existing containerised application, not rebuild it
- corporate single sign-on via the customer's Azure AD
- a small, deliberately scoped engagement – no over-engineering
The architecture
The anomaly-tracker application – first built and containerised as an MVP – was deployed to Exoscale, the Swiss public cloud, on a three-node SKS (Scalable Kubernetes Service) cluster. The whole environment is defined as code with OpenTofu/Terraform – cluster, managed PostgreSQL with automated backups, S3-compatible object storage, block storage for persistence, and a load balancer with Let's Encrypt TLS – with tfsec/checkov scanning the infrastructure code. Delivery runs on GitLab CI/CD: build, test, and containerise into the GitLab registry, then ArgoCD promotes each image into the cluster the GitOps way, with rollback. Sign-on integrates with the customer's Azure AD via MSAL.
The outcome
Delivered as the production phase of the GuardHEX anomaly-tracker engagement.
It runs in production today – built with Hydro's engineers and handed over for their own team to operate, with us on call for as long as they want.
- the anomaly tracker live on sovereign Swiss cloud – data under European jurisdiction
- every deployment reproducible from code – the environment is Terraform, not clicks
- a GitOps pipeline the team can watch, promote, and roll back on their own
- corporate SSO through the customer's existing Azure AD
- a hands-on Kubernetes/SKS workshop so Hydro's engineers can operate it themselves
- the production deployment delivered inside a short, fixed engagement
It runs in production today – on Exoscale in Switzerland, with the pipeline and the know-how handed over to Hydro's own team.