Architecture and technical due diligence
Before anything is built, a written plan: the domain model, the service boundaries, the trade-offs and the costs. You own it whether or not we build it.
Services — 01
Web application development for companies that will still be running it in five years
Multi-tenant SaaS, internal tools, CRMs, billing and the integrations underneath. We have taken a platform from an empty repository to production use by major telecom operators, and stayed long enough to live with every decision.
Typically
From €15,000 for a full build · €2,500 discovery
The problem
Most custom platforms fail slowly rather than dramatically. The architecture is chosen in the first fortnight by whoever is available, the shortcuts are reasonable at the time, and two years later every change costs three times what it should. Nobody can point at the moment it went wrong.
We have been on both sides of that. We led the Acordo project for nearly three years — a cloud-native microservices CRM for major telecom operators, designed from the first commit and then maintained long enough to find out which decisions were right. That is a different education from launching and leaving.
What’s included
Everything below, as standard.
Cloud-native service design
Independently deployable services with their own datastores and pipelines, an API gateway for ingress, and event-driven messaging only where it earns its complexity.
Data modelling that holds
PostgreSQL, MySQL or MongoDB with real constraints rather than application-level hope. Most expensive platform bugs are schema decisions wearing a disguise.
Auth, roles and billing
Authentication, authorisation, multi-tenancy, audit trails and subscription billing — the parts every platform needs and nobody scopes properly.
Observability from day one
Centralised logging, tracing and alerting. A distributed system without observability is just a system you cannot debug.
CI/CD and infrastructure as code
Terraform, Docker, Kubernetes and GitHub Actions, so your infrastructure can be recreated from source rather than remembered.
Sound like the right fit?
You probably need this if
Any of these sound familiar.
01Every change to your system takes longer than the last one
02A release requires several people and a late evening
03You are being quoted eighteen months for a rewrite
04Nobody currently on the team chose the architecture
Typical stack
TypeScriptNode.jsNestJSNext.jsReactPostgreSQLMySQLMongoDBGCPAWSTerraformKubernetes
Where we’ve done it
Work of exactly this kind.
Questions
Can you work on our existing platform?
Often, yes. We start with a short technical due diligence — usually two weeks — that produces an honest map of what you have, what is genuinely dangerous, and a ranked plan. Roughly a third of our work is inherited codebases, and a rewrite is usually the wrong answer.
How long does a platform build take?
Discovery is a week. Design is two to four. Build is typically six to sixteen weeks depending on scope, delivered in two-week increments that each end on an environment you can click through.
What happens when you leave?
Documentation, a handover, and your own team able to ship. Then either a maintenance retainer or a clean exit — both are fine by us, and the code was always yours.
Other services
Let’s build something that lasts.
Tell us about the project. If we’re not the right studio for it, we’ll say so — and usually point you at someone who is.