Idea and boundaries
We define a useful user journey, the minimum scope of the first version and the criteria for deciding whether it is ready.
Test sooner.
Choose better.
We use AI while building websites and products to explore options, create prototypes and prepare for launch. Understanding the task and taking responsibility for the result remain our work.
When it helpsDoes this sound
Does this sound
like your task?
You want to test an idea, build a first product version or use AI in development while retaining control over the logic and quality.
Scope of work
What’s included.
Prototype and iterations
We use AI for drafts, code and exploring options where it helps. We review and refine the result by hand.
Preparing for launch
We check content, core journeys and technical limitations. We distinguish an idea demonstration from a product ready for real use.
Beyond the process.
A clear outcome.
A prototype or first version with an agreed feature set and a clear development plan. We do not promise a complex product at the push of a button.
Is this a chatbot development service?
Not exclusively, or primarily. Here, AI is a tool in our website and product work. If the product itself needs an AI feature, we discuss its value, data, cost and limitations separately.
How do you estimate cost and timing?
We first discuss the task, source material and constraints. Then we agree on scope, stages, an estimate and the approval process. If the scope changes, we discuss the effect on budget and schedule before proceeding.
Which tasks do you use AI for?
Exploring options, prototyping, preparing text and visual material, and writing and reviewing parts of the code. We choose tools for the task, then review and refine their output rather than accepting it automatically.
Will the entire website be AI-generated?
That is not our objective. The priority is a website that does its job and can be maintained. AI helps at particular stages; structure, decisions, content review and launch preparation remain part of our work.
Does AI always reduce cost and time?
Not always. The benefit depends on the task, source material and required accuracy. Estimates include review, integration and refinement as well as generation. We discuss the expected value before choosing an approach.
Do we need to share internal company data with AI tools?
Many tasks can use anonymised examples or public material. If internal data is needed, we first agree on the data, the service and the permitted processing method. Confidential material does not need to be shared without that agreement.
Can the first version be developed further?
We account for this when choosing the technical foundation. Before starting, we establish whether we are making a temporary prototype to test an idea or the basis of a working product. For a prototype, we identify what would need reworking before a full launch.