Task breakdown
We look at what already runs, where the bottleneck is and what you would count as a result. We separate what would be nice from what would pay off.
01 / DESIGN
AI solution design
We work out the economics before the first line of code. Half of the AI requests we see are solved simpler and cheaper, and that is best known at the start.
Most failed AI projects did not fail in development. They failed the moment a solution was chosen before the problem was understood: a fashionable stack, a good demo, and then operating costs at real volume that outran the benefit.
We start from the other end. First we calculate what one request costs and what the whole system costs per month at your load. Then we decide whether AI belongs here at all.
Solution architecture
A diagram of components that can be maintained: where the model sits, where plain code does the job, where the database is, and what happens when each part fails.
Cost calculation
Cost per request and per month of operation at your load. Development is priced separately from running costs.
A straight answer
If the task is solved by a procedure, a script or plain automation, we say so and explain why.
What this is based on
We run our own gateway to language models with cost-based routing: some tasks go to cheap models, some to local ones, and only quality-sensitive tasks reach the expensive ones. That split is a design decision, made before any code, and it saves the largest share of the budget.
What you get
- A solution diagram that both a developer and a business owner can read
- A budget: development and operation priced separately
- A prioritised task list with an estimated effect for each item
- A clear view of where AI is not needed
When it is not needed
- If you need a two-day prototype just to look at, design work is overkill
- If the solution is already chosen and cannot change, there is nothing to calculate
If the task is not for us, we say so in the first conversation and, where we can, point you to someone who fits.
Questions we get asked
How long does design take?
From a few days to two weeks, depending on how complex the process is and how many systems we need to look at. The task breakdown and a first estimate usually fit in a week.
What if the conclusion is that AI is not needed?
Then that is what we tell you. It is a normal outcome of design work, and it is cheaper than six months of building something that never pays back.
Can we order design only, without development?
Yes. The diagram and the budget are yours, and any team can implement them. That is why we write the documentation so that someone else can pick it up.
Other services
02 / DEV AI Development We build systems that survive a provider outage, a load spike and a model change. Monitoring ships with the release, and every change has a rollback plan. Read more 03 / CONSULTING AI, Software & IT Consulting We work out why a deployed system is not delivering. Most often the model is fine and the problem sits in what gets measured, and where. The same discipline applies to software and IT decisions in general. Read more
Tell us about the task
We will look at where AI pays off and where it would cost more than it returns.