Welcome to my world
I architect scalable Data Pipelines, OCR Systems, and Secure AI Infrastructure. Specializing in
Python, Docker, and On-Premise solutions for regulated industries.
services
End-to-end design and deployment of production-grade data pipelines and AI systems, on-premise, containerized, and built to enterprise standards.
I architect and deploy AI systems that never leave your network. Full control over your models, your data, and your compliance posture, no cloud dependencies.
Turning raw customer signals, reviews, emails, voice, into structured business intelligence. Automated pipelines that surface what your customers are actually telling you.
COMMUNITY
Track Record
Bruno Bader GmbH
Architected an end-to-end Contract Management System utilizing OCR and RAG pipelines to process unstructured legal data. Built a scalable Customer Intelligence Platform (STT/NLP) using Dockerized microservices and Airflow, transforming raw audio into actionable business insights.
FZI RESEARCH CENTER
Developed Deep Learning models for Anomaly Detection on critical medical IoT sensor data. Focused on identifying malicious patterns (channel shifts, attacks) in high-stakes time-series environments.
Darkdefense GmbH
Automated critical security auditing and ransomware risk assessments via Python, increasing reporting efficiency by 90%. Hardened enterprise infrastructure against active threats
Deep Art Effects GmbH
Successfully deployed proprietary ML models to replace costly external API dependencies. Optimized PyTorch models for production image processing.
FZI Research Center
Implemented instance-based segmentation to anonymize traffic data (GDPR compliance), improving accuracy from 78% to 99%. Enabled the risk-free utilization of high-quality datasets for research by ensuring strict data privacy standards.
KIT (Karlsruhe Institute of Technology)
Focus on Machine Learning, Computer Vision, and Data Science. Research assistant at FZI Research Center.
AI ENGINEERING & OCR
Architected a high-throughput pipeline processing 1M+ legacy documents on-premise. Utilized Dockerized Python workers and Local LLMs to ensure 100% data residency while reducing manual data entry by 90%.
Tech Stack: Python | Docker | Airflow | PostgreSQL
Whether you’re evaluating an on-premise AI rollout, optimizing an existing data pipeline, or need an expert second opinion, I’m available for select projects and consulting engagements.
© 2023. All rights reserved by Leon Sander
