01
Machine Learning Engineering
ML pipelines that make it out of the notebook: from feature engineering to live, monitored systems making real decisions in production.
Berlin, Germany
Machine Learning &
Systems Engineer.
Forecasting, ERP-connected tooling and ML pipelines running daily inside a scaling Berlin food-tech operation — from research notebook to the production systems the business runs on.
I started out in scientific research and ended up building the systems a business runs on — machine learning, software, data infrastructure, and the platforms in between.
What I do
I tackle the gap between machine learning, software, and how a business actually runs.
02
Applied Data Science
Forecasting, optimization, and analytics that pay for themselves: less waste, better planning, more profit.
03
Software & Backend Systems
Backend services and REST APIs (Quarkus, Laravel, PostgreSQL, Supabase, Keycloak), React front-ends, ERP integrations, and internal platforms — the software a business runs on day to day.
04
Cloud & Infrastructure
AWS, Docker, and Cloudflare (Pages, Workers, Durable Objects) — environments you can rebuild from scratch and deployments that don't surprise you.
Machine learning, made visible
The concepts I work with — rendered live in your browser, not screenshotted.
Every panel is real, animated SVG and canvas. They pause offscreen and respect reduced motion.
Supervised learning
Labeled vs unlabeled data
Unlabeledno targets yet
Forecasting
Demand forecast · 30-day horizon
actual forecast95% interval
Live operations
Impact dashboard
- Food waste
- −0%
- Rows processed / day
- 0.0M
0%
accuracy
Weekly throughput↑ trending
Model training
Loss converging
train validationloss 0.071
Neural network
Forward pass
inputhidden ×2output
Machine LearningForecastingOptimizationData EngineeringReactTypeScriptPostgreSQLSupabaseERP SystemsAWSDockerCloudflareInternal PlatformsAnalytics Infrastructure
Selected work
From ML systems to a playable game.
Real work inside a scaling food-tech business, plus two games built from scratch that show the full-stack range.
All projectsFeatured · Playable in your browser
Child of Anon
Platformers fascinated me as a kid the way machine learning does now. So I took a game I prototyped in Python when I was just starting out and revived it with today's tooling: an original auto-scroll climber/runner in TypeScript on Phaser 3 — combat, spells, mana, real pixel parallax — deployed as its own Cloudflare Pages app. A long-time goal, finally out there.
X-STRANDEDA derelict ship, a hive that hunts by sound, and one flashlight. My own WebGL2 engine, playable in the browser.Enter the dark Case studyDemand Forecasting & Food-Waste Reduction
Forecasting models and data pipelines that sharpen demand prediction across a scaling catering operation — cutting food waste and improving resource planning.
- Forecasting
- Python
- Operational Analytics
ERP Integration & Internal Platforms
ERP development plus several internal and customer-facing web apps — connecting operational data to the people who act on it, with automation that removed manual work.
- ERP
- REST APIs
- Internal Tooling
Have a problem worth solving end-to-end?
From ML and forecasting to the custom software, ERP integrations and infrastructure around them. You don't have to solve it alone — let's talk.