George Flengas

Embedded Software Engineer & AI/ML Enthusiast

Hello! I'm George Flengas, a passionate Embedded Systems Engineer with expertise in AI/ML and hardware development. I love exploring the intersection of software and hardware to create innovative solutions, from optimizing IoT devices to training custom LLMs. With a strong background in embedded systems and machine learning, I'm constantly pushing boundaries - whether it's accelerating neural networks on FPGAs or developing intelligent IoT solutions. I thrive on learning new technologies and using them to create impactful solutions that make a difference.

Education

MEng in Electrical and Computer Engineering

Technical University of Crete, Chania, 2014-2021

Work Experience

Embedded Software Engineer

Cicicom Ltd, 7/2022-11/2023

  • Integrated radar into Gen 2 parking spot sensor, improving accuracy from 91% to nearly 100% and optimizing battery life; comparison with Bosch and Fleximodo equivalents demonstrated higher accuracy.
  • Utilized STM32 MCU ecosystem on a PCB with I2C, SPI, and UART communications, featuring BLE, electromagnetic, and LoRaWan / NBIOT communication modules.
  • Streamlined verification and setup firmware for parking sensor PCB, reducing process time from 5 to 1 minute, significantly saving factory production time and reducing associated costs.
  • Developed a GUI application for SSH communication with Raspberry Pi to automate PCB flashing and testing, accelerating the assembly line process.
  • Contributed to Gen 3 PCB development, adding new features and improving component selection.
  • Established DevOps pipelines with Jenkins and SonarCloud, enhancing workflow and code quality.
  • Developed a LoRaWAN data analytics tool using Pandas, NumPy, and Matplotlib, processing over 1 million payloads from 300+ sensors. This tool was instrumental in securing a major project by identifying critical issues and is now a cornerstone in the company’s monitoring process for all new projects, ensuring ongoing success and reliability.
  • Led testing for Gen 2 sensor, creating scripts for automated data processing and boosting productivity.

Research Assistant

MHL, Technical University of Crete, 2/2021-8/2024

  • Accelerated Convolutional Neural Networks training using FPGAs(Zynq UltraScale+ MPSoC ZCU102). Developed a Tensorflow model, replicated it in C++, and integrated into Vitis HLS for Vivado IPs
  • Achieved 1.5x acceleration over CPU, outperforming GPUs (7.8x) and CPUs (16.55x), demonstrating significant improvements in machine learning inference on FPGAs. Paper submitted to IEEE.

Skills

Programming Languages

C
C++
Python
Java
VHDL
Assembly MIPS
Matlab
PostgreSQL
MongoDB

Software Skills

Machine Learning
Data Analysis
UML
Design Patterns
Unit Testing (Google Test)
Git
TensorFlow
PyTorch
GUI Development (Tkinter)
Pandas
NumPy
Matplotlib
Flask
CMake
Makefile
GDB
Valgrind
MQTT
TCP/IP Stack
SSH
CI/CD (Jenkins)
SonarCloud
Agile
Jira

Hardware Skills

Embedded Systems
Xilinx FPGA
Microcontrollers
STM32
ESP32
AVR
Linux
Raspberry Pi
Orange Pi (Rockchip CPU)
I2C
SPI
UART
Bluetooth Low Energy
WiFi
LoRaWan
NBIOT
Circuit Design
Altium
Digital Signal Processing
Power Management Design
Oscilloscopes
Logic Analyzers

Languages

English - C2
German - B1
Greek - Native

Contact Me