The lab

KARPINITY was founded in 2015 with two purposes: to drive innovation for our customers, and to bring our own innovations to market.

KARPINITY · innovation infinity. Innovation is a loop that never closes. What we learn in client work goes into our products, and what we learn shipping our products goes back into client work. The Method runs the same way: stage 06, run and evolve, feeds back into stage 01.

The entity

KARPINITY is the trading name of Davidovski GmbH, Wehntalerstrasse 283A, 8046 Zurich, Switzerland, UID CHE-483.604.189. The Impressum holds the statutory details.

A lab with two engines

For clients. We run the KARPINITY Method: six stages from the business case to a system that is operated and reviewed every quarter. Each stage ends in a go or a stop. Three fixed-scope offers let you start small: the AI opportunity assessment, the proof of value sprint and the data readiness review.

Proven in our own products. Each method we use for clients was proven over years of designing, building and shipping products of our own. Results are on products and work.

Four hubs

We work from Zurich (headquarters), Yerevan, Nis and Skopje.

• The right specialist. Each project is staffed with the skills it needs: evaluation engineering, computer vision, on-device inference, regulatory-aware system design.
• Overlapping days. Zurich, Nis and Skopje share Central European time. Yerevan is two hours ahead in summer and three in winter, so working hours overlap for most of the day.
• One contract. You sign with Davidovski GmbH in Zurich, whichever hub your specialists sit in.
• Your data stays where you decide. We agree where data may be processed per engagement, regardless of where an engineer works. The options are on the principles page.

How we work as a lab

• Fixed scope, published duration. Each entry offer lists its deliverables, what it excludes and what we need from you. Price is on request.
• The decision before the build. We write the evaluation before the prototype, and the pilot follows the production build, so the most expensive decision rests on evidence.
• The honest no. If a pilot did not move its metric, we say so and do not recommend scaling it.
• Thought leadership. We help industry leaders apply AI and innovation with confidence, through the field notes and papers on Insights.
• Claims with sources. Every statistic traces to a primary source, labelled Swiss or global.

How we handle data and security on an engagement is on the principles page.

Start a project when you have an idea you want priced.