Björn J. Scholz, PhD

Full Stack Data Scientist / AI Engineer with a PhD in Physics ('17), specializing in production-grade AI/ML solutions. Years of experience in leading and mentoring data science teams, designing and building scalable AI platforms, and delivering end-to-end AI applications. Knowledgable in both traditional AI methods, i.e., machine and deep learning, as well as generative AI.

Bjorn J. Scholz

Skills

I lead end-to-end AI projects and feasibility studies, with hands-on experience across the complete AI lifecycle: ideation, data engineering, model development, deployment, and monitoring. My broad technical expertise enables me to deliver scalable, production-ready solutions.
Programming
Data Engineering
Data Science
Gen-/AgenticAI
Dev-/MLOps
Cloud & Big Data
Agile Methods

Public Projects

Stratosphere

Self-Hosting

A self-hosting solution for rapidly deploying curated services on a VPS. Built with Docker using Traefik as reverse proxy and CrowdSec-based security. Supports 10+ services including knowledge management (Bookstack, Docmost), project management (Planka, Kanboard), personal finance (Ghostfolio, Firefly III, Actual), RSS aggregation (Fusion), blogging (Ghost), and wikis (DokuWiki).

Stencilizer

Laser-Cutting

A tool for converting images into stencil-ready vector graphics. Uses a two-step process: first reduces the image to a user-defined color palette, then vectorizes the result into SVG format. Automates approximately 80% of the stencil creation workflow, with minor post-processing in Inkscape or similar software recommended for final refinement.

Codon Optimizer

Science

A bioinformatics tool for optimizing DNA sequences based on species-specific codon usage preferences. Accepts raw DNA or amino acid sequences as input and generates optimized sequences tailored to the target organism's codon bias. Supports configurable intron insertion for enhanced expression. Currently implements codon tables for P. pacificus (a free-living nematode), with an extensible architecture for adding additional species.

Box Forge

Laser-Cutting

A parametric box generator for laser cutting projects. Creates ready-to-cut drawing files for rectangular boxes with customizable finger joints, supporting both lidded and open designs. Lid options include hinged or lift-off styles. Configure material thickness, tooth dimensions, kerf compensation, and box dimensions to match your specifications. An integrated artwork scaling feature automatically prepares images for engraving at the correct dimensions.

Python Projects

basic-report

Reporting

A lightweight Python package for generating clean, professional static HTML reports with zero external dependencies. Built using elements of Bootstrap 4 (MIT licensed) and datatables (MIT licensed), it empowers you to create fully self-contained sites that can be served directly by any web server. The focus is set on simplicity - no oversized web framework, no database, no setup required.

Design any reports entirely in Python using an intuitive top-to-bottom workflow. Simply add elements in the order they should appear, and nest them naturally, e.g., tabs within columns, columns within tabs, or any combination you need. The API follows the logical flow of your document, making more complex layouts straightforward to build.

Perfect for data scientists, analysts, and developers who need to create shareable, standalone reports without the overhead of modern web frameworks.

Current Services


Scholz & Scholz

Website
Docmost
Planka

Riffpiraten

Website
Blog

Scholz-Lab

Website
Docmost
Planka
DokuWiki
Impressum