Most system projects start with a document. Someone writes down the requirements, someone else prices them, and a few weeks later a quote arrives for something nobody has seen. The decision is made on text and a price tag.
We do the opposite. Before you decide anything, we build a clickable prototype of the most important workflow, in five days, free of charge. You test it with your team, and only once you have seen it do we set scope and price. Here is what the week looks like.
Before the week starts
We need three things from you to make the week a good one.
- A person who knows the business. Not necessarily the manager, but the person who knows how the work actually gets done, including the exceptions and the shortcuts. That person needs to set aside about two hours on day one and half an hour on one of the following days.
- A workflow to start with. We do not build the whole system in five days. We build the most important workflow, the one with the most friction or where the most time disappears. If you are unsure which one, we help you choose on day one.
- Real-world examples. A couple of real orders, tickets or applications, ideally the spreadsheet you work in today. The prototype gets better when it is built on your actual terms rather than made-up ones.
You do not need a requirements specification. The prototype becomes the requirements specification.
Day 1: discovery call
The first day is a conversation, usually two hours, where we walk through the business, the processes and the biggest bottlenecks. We want to understand what happens from the moment something comes in until it is done: who does what, which systems are involved, where things are copied by hand and where they go wrong.
During the call we pin down the workflow the prototype should show, and what determines whether it is a success. You leave with a clear plan and shared goals for the week. The call is free, and if you want to stop there, you can.
Days 2 and 3: the build begins
This is where things that used to take weeks happen. With AI as a parallel engine in development, we set up the data model, the screens and the logic for the workflow we chose. We use your terms, your statuses and your examples, so what you see feels like yours and not like a generic template.
You do not need to do anything during these days. We may reach out with one or two quick questions when something is unclear, such as how an exception should be handled or what a field actually means. That usually takes you a few minutes.
Day 4: you test it
On day four the prototype is in your hands. It is clickable and works like a real system for the workflow we chose: you can enter a ticket, move it through the steps and see what happens.
We want you to test it with the people who will use the system, not just the person who ordered it. This is when the important things surface: that a step is missing, that a term is called something else in reality, or that what you thought was the problem was not. Collect the feedback, ideally in a half-hour call with us.
Day 5: adjustments and walkthrough
On the last day we adjust the prototype based on what you saw and walk through it together. You also get our assessment of what a complete system would include, which integrations are needed with, for example, Visma Net or monday.com, and how long it takes to get the first version into production.
After that, the decision is yours. If you want to move forward, we set scope and a fixed price before the build starts. If you do not, the week costs nothing, and you have a clickable picture of your own workflow that is worth more than most requirements documents.
What you have in hand after five days
- A clickable prototype of the most important workflow, built on your real terms and examples.
- Feedback from the team, collected before anything has been built for real.
- A clear picture of what a complete system includes and how it connects to your existing systems.
- A fixed price and a timeline, if you want to move forward.
What happens next
If you say yes, we build the full system in stages, with you continuously in the loop. The first version is in production in four to six weeks, not six to twelve months. The prototype is the starting point, so none of the work from the week is lost.
You own the code, the data and the system from day one, with no per-user license fee and no lock-in. When the business changes, the system changes with it, in short iterations as part of ongoing management. How it is possible to build this fast without losing quality is explained in our article on AI-accelerated system development.
Why we work this way
We have seen too many system projects where the decision was made on a document and the disappointment came at delivery. The prototype moves that moment to day four, when it has still not cost you anything and everything can be changed.
It also changes the conversation between us. Instead of negotiating over a list of requirements, we talk about something we can both point to. That tends to lead to better systems and fewer surprises.
If you are still unsure whether a custom system is the right path at all, we have written about the choice between custom systems and off-the-shelf systems. In practice, prototype week is the fastest way to get an answer.
Frequently asked questions
Is the prototype really free?
Yes. The discovery call and the prototype cost nothing and you do not commit to anything. Only if you choose to move forward do we set scope and a fixed price for the build.
How much time do we need to put in during the week?
About two hours on day one, a few minutes for quick questions on days two and three, and half an hour to an hour for testing and feedback on days four and five. Around four hours in total for the person who knows the business.
What if the prototype shows that we do not need a custom system?
Then we tell you. Sometimes the week shows that the problem lies in the way of working, or that an existing system can be used better. You still get a clear picture of the workflow, and we would rather have a customer who makes the right decision than a project that should never have been built.
Can we use the prototype if we choose another vendor?
The prototype is a basis for a decision, and what you learned during the week is yours. If you decide to build with someone else, you have a far better picture of your requirements than most people start with.
What types of systems are a good fit for a prototype week?
Most. Order systems, customer portals, planning and scheduling, document workflows, integrations and systems that replace spreadsheets are the most common. If yours is something else, it is usually a variant of these, and the prototype shows that in five days.
How to move forward
Book a discovery call and we will schedule day one, or read more about how we work with custom systems.