New AI-native digital products
From the first idea to daily operation, one partner. We shape the idea with you, whether it is a new digital service, a new revenue stream or a change to the business model, then design, build and run it: platforms, mobile and web apps, powered by agentic engineering.
What you get
- A business case with one named metric, an owner and today’s baseline
- A prototype measured against an evaluation written before it was built
- A production system, piloted with users in one place
- The product extended across the scope that earned it
- Operation under a written agreement, with a value review every quarter
Who it is for
You see room for a new digital product, a new revenue stream or a change to how your business earns, and AI is part of it. Sometimes the opportunity is clear and its shape is not yet. The usual route passes it through four suppliers: strategy, design, engineering and operations. Each handover loses context, and nobody owns the result.
We shape the idea with you, then take it the whole way. The lab that writes your business case also builds the prototype, engineers the product, pilots it and runs it.
What we build
New digital products and services, built with agentic engineering:
- Digital services that open a new revenue stream, designed down to the business model.
- Customer apps on iOS, Android and the web.
- Agents on messaging channels and in the back office.
- Internal platforms that connect to the systems you already run.
- The operating layer: monitoring, cost per completed task, regression tests and a value review.
Most products draw on one or more of our other services: agentic platforms, AI customer experience or connected intelligence.
How an engagement runs
The work follows the six stages of the KARPINITY Method.
| Stage | What you get | Duration |
|---|---|---|
| 01 · Ground the case | A business case with one metric, an owner and a baseline, and a verdict on the data | 3-5 weeks |
| 02 · Prove the value | A prototype and an evaluation written before it was built | 3-6 weeks |
| 03 · Engineer the system | A production system scoped to the pilot’s footprint | 8-16 weeks |
| 04 · Pilot in one place | Users in one team or site, and the metric measured | 6-12 weeks |
| 05 · Scale what works | The product extended across the agreed scope, value measured again | 3-6 months |
| 06 · Run and evolve | Operation, monitoring and a quarterly value review | Ongoing |
Every stage ends with a written recommendation and your decision: continue, adjust or stop. Each stage is contracted separately, and the price is quoted on request once its scope is fixed.
The pilot comes after the production build. Users and their personal data should not meet a system before its security, data protection and human oversight exist. So stage 03 builds for one site only, and stage 05 does the engineering that wider use needs.
After go-live we run the product under an operating agreement with defined service levels. If you would rather run it yourselves, we hand over the runbooks, the evaluation suite and the dashboards, and support the handover period.
Built for Switzerland
Where the product runs is settled in stage 01: Swiss-hosted cloud, EU-hosted cloud, your own infrastructure, on premises or on the device. Every engagement that touches personal data gets a data protection impact assessment, early enough to change the design. If the product’s output reaches the EU, stage 01 also screens it against the EU AI Act. The rules, with dates and sources, are on the regulatory reference.
Proof from the lab
Each method here comes from designing, building and shipping products of our own.
| Build | Label | What it shows |
|---|---|---|
| Nauvo | In the App Store | A ninety-second naval duel, taken from idea to the App Store, Google Play and the browser |
| Telebimmer | Live | A Telegram service for BMW owners, built on the official BMW CarData service. Not affiliated with BMW AG |
| Zeza | Beta | A voice app with speech, writing help and read-back, all on the phone |
| Kutio | Beta | A Mac archiver with no accounts, no analytics and no network code |
| A multi-channel AI concierge | In production | Five customer channels, cost-aware routing across model providers, a staff dashboard with cost per channel |
| A helm for a fully electric boat | Client delivery | A first working prototype, from a blank page to a demonstration on the water in Spain, on the client’s own hardware and to the client’s deadline |
Built on
Questions
Do we need to arrive with a finished idea?
No. We shape it with you: where a new digital service, a new revenue stream or a change to the business model could pay off, and which version is worth building first. Stage 01 turns it into a business case with one named metric, an owner and a baseline. The AI Opportunity Assessment does that in 1-2 weeks.
Do we have to commit to all six stages?
No. Each stage is contracted on its own, and its price is quoted once its scope is fixed. You can stop after any stage with nothing further owed.
How long until users have the product?
At their full published ranges, stages 01 to 03 take 14 to 27 weeks, and the pilot starts after that. Starting with one of the fixed-scope offers shortens the first two stages.
Who runs the product after launch?
We run it with you under the Run and evolve agreement, or your team takes it over with our runbooks, evaluation suite, persona cards and dashboards. You can change that choice at any quarterly review.
What if the idea does not hold up?
Then we say so in writing at the next decision gate, with the evidence. For example: the pilot did not move the metric, so we are not scaling it. A gate is where stopping costs least.
Next step
Start with the business case
Start with the AI Opportunity Assessment: in 1-2 weeks your idea becomes a business case with a named metric and owner, a regulatory applicability note, a first read on the data and a recommendation. The price is on request.
See the AI Opportunity Assessment