Independent contractorPoznań, PolandAvailable for engagements

Embedded software for devices where failure is not an option.

Firmware, embedded Linux and the engineering process around them. Over twenty-five years in medical diagnostics, road and tunnel safety, industrial automation and IoT, from the first line of code to the team that maintains it.

Start a conversation Read the case studies

Portrait of Radosław Gołębiewski — Radosław Gołębiewski, PhD Principal Embedded Software Engineer

25+: years in embedded systems
SIL 2 / 3: safety systems delivered in UK road infrastructure
Class D: IVDR medical diagnostics, the highest-risk category
0 → 20+: engineers: a team and its process built from scratch

01 — What I do

The whole device stack, and the process that keeps it shippable.

Engagements range from a few days of review to multi-year ownership of a product's software. Remote or on site.

Embedded software and firmware

C/C++ on ARM Cortex-A/M, ESP, Microchip and TI; bare metal, FreeRTOS, Zephyr.

Embedded Linux

Yocto images, drivers, systemd services, board bring-up, build and release pipelines.

Safety-related and regulated systems

Software and hardware developed to SIL 2 / SIL 3 and for regulated medical devices.

Software engineering process

Version control, code review, CI/CD, unit, integration and system testing, packaging and release; set up for teams from zero to 20+ engineers.

Device data and tooling

Data acquisition, remote device management, test and diagnostic tools in C++ and Python.

AI-assisted development

Specification-driven, test-first delivery with AI agents, independent automated reviews and a human in charge of every decision.

02 — Experience

Four industries where software errors have physical consequences.

Medical diagnostics: Principal software engineer for a next-generation DNA sequencing platform (IVD Class D under EU IVDR, Class II/III under FDA, the highest-risk category): chip data acquisition, sequencing pipeline tools, experiment data management, the first Yocto image of the device.
Road and tunnel safety: Over ten years of real-time highway control systems in the UK: the country's first SIL 2 tunnel emergency management system, a SIL 3 network changeover unit, roadside controllers, Bluetooth traffic monitoring and camera enforcement systems.
IoT and connected devices: Co-founded WizzDev, an embedded software house, grew its engineering team from zero to 20+ and established its agile workflow, CI/CD and testing practices; delivered data acquisition, remote management and predictive maintenance for industrial, defence and smart home clients.
Industrial automation and measurement: Automation projects for international manufacturers and technical consulting in process measurement.
Research and teaching: MSc in Electronics and Telecommunications and a PhD in telecommunications, specialising in Digital Signal Processing (DSP); formerly a lecturer in electronics at Poznań University of Technology.

C / C++
Python
Yocto
FreeRTOS
Zephyr
ARM Cortex-A/M
ESP
CI/CD
DSP

Full history on LinkedIn

03 — Case studies

Two devices, end to end: firmware, Linux, application and the pipeline that ships them.

Automotive diagnostics · Product

CAN Bus Analyser: a diagnostic device from concept to product

A portable analyser for automotive diagnostics, developed at my company. It records CAN traffic together with the analogue voltage levels on the bus, so faults that never show up in the digital data can be found, even in a moving vehicle and remotely.

My role: led the project and wrote most of the firmware, the Linux image and the desktop application.

Interface: STM32F407 firmware on FreeRTOS: CAN, bus voltage sampling at up to 3 MSa/s, USB.
Device: An embedded Linux (Yocto) image with a web interface for remote configuration and statistics over Wi-Fi.
Desktop: A C++/Qt analysis application for Linux and Windows, built and packaged in CI.

IoT · AI-driven lifecycle · Personal project

An IoT watering system built with an AI-driven lifecycle

My capstone project for the 10xDevs programme in AI-native software engineering: a per-pot automatic watering device that waters on its own and reports what it actually did, with a hard safety stop so the pump never runs dry.

Built after hours and verified on real hardware. Every decision mine.

Device: MicroPython firmware on the Raspberry Pi Pico 2 W and ESP32-C6.
Service: A Django web app on a low-cost Rockchip Linux board, running my own Yocto image with reproducible builds, a firewall and OTA updates with automatic rollback.
Method: Requirements, architecture and plan first; test-first implementation by AI agents, an independent automated review of every change and CI with end-to-end tests.

04 — How I work

Faster with AI. As rigorous as ever.

I now use AI-assisted development to deliver faster, while keeping testing and review as rigorous as ever: specification first, tests before code, an independent automated review of every change, and a human in charge of every decision.

Engagement: As an independent contractor through my company: remote or on site, short consultations or long engagements.
Clients: In the UK, Germany, Poland and across Europe.
Languages: English and Polish; basic German and Spanish.

05 — Contact

Have a device, a deadline, or a team that needs a process? Write to me.

A short description of the system and where it stands is enough to start.

Email: radoslaw.golebiewski@rj2.pl
Phone: +48 515 150 043
LinkedIn: radoslaw-golebiewski-rj2
Based in: Poznań, Poland