Introduction
Hello and welcome to my curriculum vitae.
I am Bart, a data architect with many years of programming experience, and it is my passion to create a Linked Data world where data is findable, accessible, interoperable and reusable (FAIR). Let me help you create, maintain and make use of ontologies to formally describe your data in a machine-readable way; to make the right architectural choices to prevent bottlenecks and rigid top-down management and empower data engineers to autonomously produce their data; to seamlessly exchange and integrate data even outside of the company; to enhance collaboration between domain experts and data engineers; to build knowledge graphs so data scientists can get the most out of their LLMs.
I am particularly familiar with the Semantic Web technology stack (RDF(S), OWL, SHACL, SPARQL, etc.) and LinkML for data and information modeling, and with Data Mesh for a bottom-up, decentralized architecture, but since I have a firm grasp of the underlying principles I can easily adapt to different technologies and paradigms. I also have experience writing software which transforms data models from one metamodel to another, e.g. transform a SHACL shapes graph to an Avro schema.
As a colleague I really like helping others. I’m patient and a careful listener. I’m not very fond of chit-chat, but if we share an interest or have a similar sense of humor you’ll probably find me a lot more talkative.
I love discussing deeper philosophical and scientific topics. Other (perhaps obvious) interests are computer science and mathematics. I’m crazy about music and film, and I also like to read. If you’re curious about my taste, please click the Music, Films and Books links in the Interests section. And to prove to you I’m not a snob: I also like cartoons (but only the good ones).
If you think there’s a nice opportunity to work together, or even if you share some of my interests and would like to chat, please feel free to drop me an e-mail. I would love to get to know you better.
Experiences
- I’ve collaborated with other architects and teams to implement a Data Mesh architecture in the organization. Thanks to my technical background, I can speak the language of both business and IT, which really helps to bridge the gap.
- I am one of the few experts in using the Semantic Web technology stack for our data and information modeling to facilitate strong and formal semantics, flexible decoupling of conceptual from logical data models and to enable code generation, etc.
- I’ve developed the Metamorph project (written in Clojure), which generates technical schemas from a logical data model in SHACL.
In this team I helped realizing image recognition solutions. We trained object detection models to recognize gas and electricity meters, their readings and other related componenets. What I did mostly was:
- Building and maintaining RESTful APIs, and reusable utility packages.
- Improving our software and development quality, i.e. CI/CD, package management and tools such as code autoformatters.
- Write scripts to transform data or that make part of a machine learning pipeline.
I was part of the team responsible for managing the IT process of contractors and mechanics that had to maintain or install customer gas and electricity installations. I was responsible for taking care of automating the testing process:
- I wrote business-friendly test suites, generic resuable test code for use in test suites (also outside the team) and libraries to extend Robot Framework’s capabilities.
- I created a set of programs, including a Windows GUI, to allow manual testing flexibly. I automated fetching data from SAP and writing it to Excel sheets. Users could adjust and complete the data there, and generate or send a SOAP message with the push of a button. I was the sole maintainer of these programs, including releasing and guiding users.
Using the Symfony framework, I developed new features for the website (on a freelance basis).
- Writing and maintaining XSL templates that crawled the web for news content.
- To improve our workflow, I turned a simple sheet into a more practical collaborative ‘todo-list-like’ system (called The Maintenance Sheet) for our developers, including the automatic import of log entries from our backend software.
- These companies provided internships, which I applied for since I was in need for a job. I did not hang around for long at these places.
- I wrote small BI applications using the rapid development tool Kettle.
- Specializing in webshops, e*cube had an impressive database setup with associated APIs. Here I learned SQL and got my first taste of programming from the MVC philosophy.
- I’ve become fluent on the Linux command-line and also wrote some shell scripts for system administation purposes.
- As the tidy perfectionist of the team, I maintained the company wiki and much of the project documentation.
- I took upon me most of the service desk tasks, supporting customers and keeping our backlog of reported issues orderly.
- Every Saturday I sorted and delivered mail.
Open Source Projects
The most noteworthy open source projects I've worked on with varying degrees of involvement.
Metamorph is a libary that enables the generation of Avro schemas from a logical data model expressed in SHACL. It uses Semantic Web technology and is written in Clojure. An earlier version of this project was written in functional style Kotlin).
This is an API I built on top of the Beets music organizer. It’s a modern alternative to its standard Web plugin, extending its capabilities as well as being built in a more modern framework (FastAPI).
An open-source Robot Framework library I wrote to work with Excel 2010 (and above) files.
A backup script I wrote for an easy configurable back-up solution using Rsync.
I contributed to this open source music organizer, providing the ability to set metadata fields when importing an album into your library.
I reverse-engineered how the music player MPD caches its audio metadata, and wrote a very fast script that generates such a cache file from a given Beets database file.
My solutions to the yearly Advent of Code puzzles. Each year has its own branch. These are complicated problems you have to solve as quickly as you can, so please don’t expect beautiful code.