I've spent the last 3 years shipping backend systems that handle real load - and I've learned that the best code is the code your future self won't hate. Right now, I'm deep in my MSc program, diving into reinforcement learning and exploring how to make AI/ML systems reliable, observable, and actually useful. I gravitate toward messy problems: unclear requirements, legacy codebases, systems that need to scale yesterday. Give me an ambiguous spec and I'll give you a stable API, a data pipeline that doesn't drop records, and a deployment process that won't wake you up at 3am. Outside of code, I'm usually hiking, making music, or figuring out how to get better at the violin (slowly).
ESTAR
Capgemini
Capgemini
Blockchain Analysis Tool
Clustered Bitcoin addresses with a multi-input heuristic and enriched entities using public data sources.
Multi-threaded Key-Value Store
Built a from-scratch object store with multi-threaded client/server architecture over Unix sockets.
On-Prem CI/CD & Release Automation
Built a compliant deployment pipeline for a hospital system with strict networking and audit constraints.
If you're building production AI/ML systems (or the backend/platform foundations under them), I'd love to talk. I'm open to roles in Copenhagen/EU and to freelance projects while studying. I write about backend architecture, reliability, and what I'm learning in reinforcement learning along the way. If you're reaching out, feel free to include a couple of links or a short description of the problem - I'll reply.