Freelance · Athens, GR · Available

Building things that think,
sense, and talk to each other.

I'm Ierotheos Katsoulis — an electrical & electronics engineer designing embedded systems, IoT hardware, and machine-to-machine protocols. From custom PCBs and J1939 vehicle diagnostics to voice assistants on Raspberry Pi, I take projects from schematic to working product.

→ Start a project View work
/ 01

About

I've spent the last several years at the intersection of hardware and software — designing PCBs, writing firmware for STM32 and ESP32, building CAN/J1939 diagnostic systems for vehicles, and deploying ML on edge devices.

Alongside freelance work I'm a service engineer at Divitec S.A., and I've taught Optoelectronics and Quantum Mechanics at the University of West Attica, where I also contributed to published research in refractometry of attenuating media.

If your problem involves a microcontroller, a sensor, a bus protocol, or anything between bare metal and the cloud — let's talk.

// Quick specs

LocationAthens, Greece
StatusAccepting work
Experience7+ years
EducationBSc EEE · UniWA
PublishedIOP · 2022
LanguagesEN · EL · FR
/ 02

Selected work

★ FEATURED CASE STUDY Workbench: digital microscope inspecting populated PCB, two boards in foreground, 3D printer behind
Custom hardware · Client product · Shipped

Electronic scoring board for fencing equipment

A custom-designed scoring board built end-to-end for a sports equipment manufacturer. Schematic capture, PCB layout, SMD assembly, microscope-assisted rework, 3D-printed enclosure, and final integration — every step done in-house. Each board ships with the AXIOSYN mark on the silkscreen.

Top-down view of the assembled custom PCB showing the QFN microcontroller, crystal, regulator, TVS diode, and AXIOSYN silkscreen branding Angled view of the populated PCB on the workbench Finished 3D-printed black enclosure with three 1/4 inch jack connectors and two cable glands
PCB design KiCad SMD assembly Firmware 3D-printed enclosure Microscope rework

J1939 remote vehicle diagnostics

Embedded · IoT

End-to-end diagnostic system for heavy vehicles: hardware interface, J1939 frame decoding, remote control logic, and a data pipeline for real-time fleet monitoring.

CAN busJ1939STM32C/C++Node.js

Edge voice assistant on Raspberry Pi

Voice AI · Edge

Local-and-cloud hybrid voice assistant running on Raspberry Pi. Optimized voice pipeline (VAD → ASR → intent) and LLM orchestration with privacy-respecting on-device fallback.

Raspberry PiPythonLLM orchestrationLinux

Facial expression recognition on embedded

Computer Vision · ML

Deployed OpenCV-based facial expression recognition on low-power embedded devices. Tuned the pipeline for real-time inference under tight memory and CPU budgets.

OpenCVPythonPyTorchEmbedded Linux

Audio electronics: amps & footswitches

Analog · PCB

Designed audio amplifiers and guitar footswitches from schematic to PCB. Also handle repair and electronic debugging of existing audio hardware.

Analog designKiCadLTSpiceSoldering

Web3 frontends & wallet auth

Web · Blockchain

Built blockchain-integrated frontends with wallet authentication flows and clean interfaces for on-chain data — making complex contract interactions feel like a normal web app.

JavaScriptHTML/CSSWallet authOn-chain data
/ 03

Stack

// Embedded

  • STM32
  • ESP32
  • Raspberry Pi
  • Arduino
  • RTOS

// Protocols

  • CAN bus / J1939
  • SPI / I2C / UART
  • Modbus

// Languages

  • C / C++
  • Python
  • JavaScript
  • C#
  • VHDL · MATLAB · SQL

// Tools

  • KiCAD
  • Proteus
  • LTSpice
  • Simulink
  • Node.js · PyTorch

// Systems

  • NixOS / Linux
  • LLM orchestration
  • OpenCV
  • 3D modeling & printing

// Research

  • Optical characterization
  • Refractometry
  • Statistical error analysis
  • Published: IOP 2022
/ 04

Get in touch

Got a project? Let's talk.

From peer-reviewed research and full product builds to a board that just needs fresh eyes — no job too complex, no fix too small.

→ ierokatsoulis@gmail.com → LinkedIn → +30 694 777 9368