Edge AI – it’s so much
deeeeeper.

AI does not have to live in a data centre. We research and develop Edge AI that brings language and vision models directly into applications — running locally on ordinary CPUs, self-contained and remarkably lean, without an army of dependencies.

Our own V7 Engine combines local inference with retrieval, full-text search and persistent memory. Our RAG Rewriting technology takes retrieval a step further, injecting relevant knowledge directly into the model’s ongoing reasoning instead of simply adding it to the prompt. Together, they open up new possibilities for applications that need intelligence without relying on the cloud.

Less infrastructure. More control. Quite a bit more fun.

Embedded Graphics
Programming

We like software that knows what the hardware is doing. We develop embedded systems, graphics and computer vision for demanding applications, including automotive — right down where code meets silicon.

From efficient low-level software to graphics pipelines and image processing, we work where performance, memory and hardware constraints actually matter. We understand that good engineering is about finding the right balance between the software, the hardware and the task at hand.

Web & App
Development

Need a database, online presence or mobile app? We build practical software for businesses, institutions and medical practices — from focused individual websites to complete applications.

We combine clean interfaces with solid technology underneath, connecting databases, APIs, business logic and user-facing applications. No five-year IT strategy, no technology for technology’s sake. We listen to your needs and ideas, find the sensible way to turn them into software, and build solutions that get the job done.

Advantages included

Have something in mind?
Let’s make it clever.

cpki experience

We have been writing software since the days when the Internet was still making dial-up noises. We started with video game and console coding, then moved into databases, operational data processing, websites and mobile apps. Graphics and panorama rendering followed. Then computer vision, machine learning and deep learning. Image signal processors. And now AI that runs locally on a CPU.

The technology kept changing, and so did the machines — but the interesting part stayed the same: finding out what they can do and then making them do it. We have followed the technology wherever it got interesting, from early pixels and processors to today’s AI systems — and occasionally built our own.