Backend and sovereign cloud
APIs and architectures that hold up under load, hosted on Swiss cloud or on your own infrastructure. With no dependency on a vendor whose pricing and jurisdiction you do not control.
Where your data lives
The question comes before the architecture. Depending on your sector, the acceptable answer ranges from 'in Switzerland, with a Swiss host' to 'in our own server room, with no outbound internet access'. We start from that constraint and design afterwards, rather than the other way round.
In practice three options cover almost every case: Swiss cloud at Infomaniak or Exoscale, hosting on your premises, or a hybrid model with the sensitive data at your site and the rest elsewhere. All three have very different consequences for running costs, and we cost them before deciding.
What we put in place
Our architectures stay deliberately simple. A system your team cannot understand is a system nobody will be able to fix at three in the morning.
APIs and business services
REST or GraphQL, documented, versioned, tested. Built to be consumed by your mobile and web applications and by your partners.
Self-hosted Supabase
PostgreSQL, authentication, storage and realtime, on your infrastructure. The productivity of a managed platform without the dependency.
Data processing
Ingestion, transformation and synchronisation pipelines between existing systems, with error recovery.
Observability
Logs, metrics, alerts. Knowing a service is degraded before a user calls.
Leaving a vendor without rewriting everything
Migrating away from an American platform is rarely a big-bang project. We proceed in stages: first isolate the dependencies behind interfaces, then move components one at a time, starting with those carrying the most sensitive data.
That approach keeps the service running throughout the migration, and allows you to stop halfway if the economics do not hold. We always cost the running expenditure before and after.
What happens after go-live
An architecture is not delivered on the day it goes live: that is the day it begins. We install monitoring, define alert thresholds, document recovery procedures and test the backups by actually restoring them.
The maintenance contract is sized on criticality, not on a standard package. For a system whose failure halts production, that may mean an on-call rota. For an internal tool, intervention during office hours is enough and costs far less.
Reversibility
You must be able to replace us. That is a written commitment, not a statement of intent.
- Infrastructure described as code, versioned in your repository.
- No proprietary MCDev component anywhere in the production chain.
- Documented operating procedures, tested with your teams.
- Complete data export in an open format, at any time and at no cost.
An example
A sovereign self-hosted backend for a field iPad application: offline synchronisation, analytics dashboard and storage of operational data on infrastructure the client controls.
Frequently asked questions
Where is the data hosted?
Wherever you decide: with a Swiss host, on your own premises, or in a hybrid model. We give you the running cost of each option before choosing, because the gap is often smaller than expected.
Is self-hosted Supabase viable in production?
Yes, provided you accept the operational work: backups, version upgrades, monitoring. That is what we set up and document. In exchange you keep PostgreSQL, a standard format and the ability to leave.
Can you take over an existing infrastructure?
Yes. We start with an audit of what exists, its costs and its risks, then propose a staged plan. We only recommend a full redesign when the audit justifies it with figures.
What availability can you guarantee?
It depends on the architecture you fund. Full redundancy is expensive and is not justified everywhere. We start from your real tolerance for interruption and size accordingly, with a written commitment.
What happens if we stop working with you?
You get everything back: the code, the infrastructure described as code, the data in an open format and the operating documentation. We support the handover to the team taking over, and it is written into the contract.
Let's talk about your project.
Describe your need in a few lines. We answer within 48 working hours, with a first technical opinion rather than a sales pitch.