Pieter Fiers

Machine Learning, Maps & Microcontrollers



From building autonomous robots, drones, and rovers, to helping us map the world with apps and websites: if it involves a computer interacting with the outdoors, I'm in. Explore my portfolio of past projects here: pfiers.net​/projects.


MSc Computer Science - Cardiff Metropolitan University, Cardiff

2021 - 2022

Magna cum laude

Bachelor Applied Informatics - UCLL, Leuven

2018 - 2021

Magna cum laude


Data Engineer - Dataroots, Leuven

March 2023 - present

As a Data & Cloud Engineer at Dataroots, I collaborate with customers from all sectors to create truly end-to-end data solutions. This includes not just crafting state-of-the-art models and analysis platforms, but also building a long term data strategy and setting up the cloud infrastructure to support it all.

Joining a team of five, I started my first project at dataroots in April of 2023. Our client, CluePoints, offers a risk and quality management platform for clinical trials, used among others by Pfizer and the FDA. The goal of our team was to eliminate bottlenecks and increase flexibility in the trial data ingestion pipeline. To this end I deployed and benchmarked multiple datalake components including the Airflow and Prefect workflow orchestrators, databricks and snowflake data intelligence platforms, and a combination of dbt, Spark, and custom Python data pipelines. We were able to reduce the storage requirements by an order of magnitude and to greatly increase the speed and flexibility of trial data transformations.

On my second and current project, I am collaborating with the audience data intelligence team of VRT, the dutch-language public broadcaster of Belgium. The goal of this project is to integrate viewership and engagement data from multiple sources into the company-wide datalake. To best support the existing workflows of the team, I have developed an Airflow-orchestrated data export task server for Windows Server. This FastAPI Python server uses a combination of accessibility automation frameworks and UI machine vision to extract data from GUI-only applications. I visualized this data in a set of comprehensive PowerBI dashboards which empower data-driven decisions. My solution saves countless hours in previously manual weekly report generation.

I also advised, and helped my teammates in various projects, for among others, C-Energy, Touring, and Immoscoop. I did this in particularly through my expertise in machine vision as well as infrastructure and system administration.

Tech stack:

Python, SQL, PowerBI, DAX, Power Query M, Spark, MLflow, Machine Learning, OpenCV, Azure, AWS, Airflow, Prefect, Docker, Numpy, databricks, Cloudformation/Terraform, dbt, Great Expectations, Agile

Professional Tutor - classgap.com and other platforms

2019 - present

I first started tutoring in 2019 for my then university, UCLL. Since then, I've taught over fifty students in computer science and software engineering topics through various organizations. You can read reviews by my students on pfiers.net/cgr.

Tech stack:

Python, SQL, JavaScript, TypeScript, React, Java, C++

Lead Blockchain Engineer - Art_Value, Stockholm


Web3 and blockchain engineer for NFT and fungible token art auction platform. As a blockchain engineer I developed and deployed solidity smart contracts for the Ethereum blockchain to facilitate the trading of auction tokens. I also developed the web3 website in Typescript with Vue.js and backend in PHP and Python with a PostgreSQL database. The backend was deployed on Google Cloud Platform using terraform. I oversaw the handover of the web app to a team of three junior full stack developers.

Tech stack:

SQL, Solidity, Typescript, PHP, Python, Terraform, Google Cloud Platform

Intern Software Developer - Open Summer of Code - osoc.be


Open Summer of Code is a 4-week summer programme focused on building open source projects of public utility. Commissioned by Brussels Mobility, my team and I built Cyclofix: A solution to map, present and integrate cycling infrastructure in Brussels.

Tech stack:

TypeScript, Node.js, Mapbox

Intern Computer Forensic Analyst - i-Force, Aalst


Student job at i-Force, a fraud and digital forensics company. During my time here I created tools to aid in data recovery and processing for servers affected by a ransomware attack. I also gained practical experience with Palo Alto Firewalls and Kerberos.

Tech stack:

Python, Citrix Hypervisor, Windows Server, Kerberos, Palo Alto Firewalls

Intern Research Software Developer - NERF (IMEC), Leuven


Internship at the Belgium neuroscience lab of professor Karl Farrow, part of NERF (Neuro-Electronics Research Flanders). During this assignment I developed “Kinect Mouse Tracker”, a computer vision application to track the position of a mouse in real-time using depth information.

Tech stack:

C++, OpenCV, Object detection, Multi-stream processing, Depth cameras


Please do check out my projects in detail with links to source code and results on pfiers.net/​projects. In addition to the two projects listed below, I've also developed Android apps, websites, and embedded systems.

Lakehopper - Deep Learning, Machine Vision, Python, Rust

For my Master’s thesis, I developed the high-level planning software for Lakehopper, an autonomous drone that can fly indefinitely by hopping from one lake to the next, recharging its batteries each time using solar panels.

This system uses a custom designed deep convolutional neural network that I trained to identify lakes and buildings in aerial imagery. It further uses OpenCV for object detection based on this data. From this, it generates a navigation graph to calculate the best multi-hop paths between lakes. These paths avoid build-up areas and restricted airspace.

In my free time I have created a first prototype version of the drone and the next generation of the planner software, including integration with an on-board single board computer running ROS, the Robot Operating System on Linux.

Pathy - Deep Learning, Machine Vision, Python, Robotics

Pathy is my Bachelor’s thesis project. It is a small tank-track rover that autonomously follows forest paths using a Tensorflow + Keras deep convolutional neural network. The network runs on the single board computer on the rover and processes data from a live webcam feed. This pipeline is facilitated and controlled through ROS. The labelling of the training data was crowdsourced to other students through a webapp.


I believe that the best way to learn is by doing, and I have a track record of quickly picking up new skills and applying them in real-world end-to-end projects.

Through professional experience, personal projects, and my bachelor and master's theses, I have gained profound knowledge and the ability to practically apply deep learning and machine vision techniques for object detection, classification, and segmentation.

Over my more than 7 years of focused software and systems development, I have mastered an extensive range of technologies and programming languages. Thanks to my computer science master's degree, I have a solid understanding of the fundamentals of the field, as well as of modern data science techniques.

Through my undergraduate specialization in systems and networks and professional experience, I also developed a deep understanding of application and network security.


When I’m not programming I’m usually either hiking, climbing, geocaching or sailing. I also enjoy teaching others what I’ve learned. I’m a certified sailing coach and I tutor others in technological topics. Many of my projects are also inspired by my love for cartography and GIS. I’m a member of OSM Belgium and the OpenStreetMap Foundation, and I’ve contributed several open-source apps and websites.