Engineering from the circuit board to the cloud

We do FPGA design in VHDL, embedded firmware in C, IoT telemetry, data pipelines and applied machine learning. Our documented work is mostly academic and R&D, and each project page says so.

Some products go beyond the browser. They read sensors, process a video stream in hardware or collect measurements from many devices. This is the lowest-volume part of our work, and we want to be clear about where it comes from.

What is the provenance of this work?

We have documented 15 engineering projects. Most are academic or R&D projects. One is an internal product and one an independent project. None is described as paid client work, and each project page says what it is.

That matters when you judge us. These pages show how we think and what we can build. They are not references from customers.

What can we do?

FPGA and digital design

We write RTL in VHDL, with testbenches, and implement it in Vivado. Our streaming image demosaicing pipeline turns a raw camera sensor stream into colour images at one pixel per clock cycle. It was checked against a Python reference, and timing closure at 100 MHz is documented. A second project builds a FIR filter datapath from a multiply-accumulate unit, memory and control logic.

Embedded firmware and IoT

The AVR telemetry project runs on an ATmega328PB at 16 MHz. It handles interrupts, ADC and PWM, I2C and 1-Wire sensors, and a UART link to an ESP8266 that sends JSON over HTTP. A separate Linux kernel project covers character devices, locking and a small filesystem.

Data and machine learning

We built a distributed geospatial analysis in Spark, implementing one analysis three ways and timing each. A remote-sensing project applies machine learning to satellite imagery. We also have notebooks on image classification and NLP, all listed on the work page.

Why do hardware and software belong together?

A device is only useful when its data reaches someone. Telemetry needs an ingestion API, a database, a dashboard and alerts. The same studio can build both ends, so the interface between them is designed once.

How would a project work?

  1. We list the real constraints: sensors, power, speed, cost per unit, certification needs.
  2. We choose the smallest design that meets them, and say what must be tested on real hardware.
  3. We build and verify against a reference, as in the projects above.

We do not claim certification, production volumes or field deployments that we cannot show.

Where to go next

If your product needs a web or cloud side as well, see software. For a hardware question, contact us.

Questions we get asked

Is this client work?

Mostly no. Of the projects we document, most are academic or R&D work. Each page states its origin. Two are not: the operations and workforce CRM is an internal product, and a marketing site with a blog is an independent project. We do not present any of it as paid client work.

Do you take on FPGA or embedded projects?

We are open to them, especially where hardware has to connect to software, data or a cloud service. Tell us the constraints and we will say honestly whether it is a good fit.

Can the hardware connect to your software services?

That is the point of offering both. A device that sends telemetry needs an API, a database and a dashboard, and we build those as well.

Can you share the source code of these projects?

The repositories are private, and the project pages do not publish them. The pages describe what was built, how it was checked and what it shows.

Tell us what you want to build.

A product, a website, a search problem or a system that should work better. We reply with a plain answer and the first step.

Email
[email protected]
Phone
+30 698 179 8240

A few sentences are enough (at least 20 characters).

We use your details only to answer your inquiry. See our privacy policy.