Custom software

We build the software your business needs when standard tools do not fit: web applications, internal tools and integrations, produced with AI assistance and reviewed by engineers.

What we solve

We build with AI assistance: it speeds up writing code, but every delivery passes engineering review, testing and agreed acceptance criteria. That describes how we build software, not whether the program includes AI features: that is decided separately for your case. Generated code is not automatically safe.

We define scope from the outcome you expect, validate early with a usable version and build a maintainable foundation that can evolve with your operation.

Common problems

  • A spreadsheet supports a critical process and is starting to break
  • The team copies data between applications that do not talk to each other
  • The current SaaS product is expensive, excessive and still misses the essentials
  • A prototype needs to become an operable product

An illustrative example

Before: a small distributor takes orders by phone and messaging, logs them in a shared sheet and prepares delivery notes by hand. Transcription errors cause returns every week.

After: a simple internal application where each order is entered once, stock is discounted and the delivery note is generated. The team still validates every order before sending it.

What the engagement includes

  • A working application, deployed and accessible
  • Acceptance criteria verified with you
  • Usage and maintenance documentation
  • A launch session with your team
  • An optional evolution and support plan

What we need from you

  • A description of the process as it works today
  • Sample data for testing, with no real credentials at first
  • One person who validates progress and decisions
  • An agreed minimum viable scope before we start

Scope and costs

Scope, milestones and price are agreed in writing before we start. We usually work per project with a minimum viable scope and later extensions where they add value.

Third-party costs (hosting, domains, licences or AI model usage) are estimated separately; costs, account ownership and the billing arrangement are agreed in the proposal. Later maintenance is optional, varies with complexity and dependencies, and is also agreed in writing.

Frequent questions

Does artificial intelligence write the code?

In part: it speeds up production, but engineers review, test and accept every delivery. Responsibility for the result stays human.

Who owns the code?

Ownership and usage licences are agreed in each project's contract, respecting third-party and open-source licences used.

How long does a project take?

It depends on scope, data migration and integrations. Each proposal includes an agreed schedule with milestones.

What if I already pay for a SaaS product that almost fits?

Sometimes integrating it beats building. We assess that with you before proposing any new development.

Keep exploring

Systems integration

Process automation

Custom software or SaaS: how to decide

What the project delivers

  • Working application
  • Integrations and API
  • Evolution and support