I build production backend systems and ML pipelines. MSc Information & Communication Systems, TU Hamburg. Based in Germany.

Applied Biomedical Systems GmbH (07/2026-Present) — Application Engineer

• Programming of components and applications for internal software development as well as for end-users (medical professionals and scientists), and providing user support for software installation and usage.

Bündnis 90 / Die Grünen, Hamburg Parliament (03/2026-05/2026) — AI Intern

• Researched and analysed policies including ethical, regulatory, and societal implications using AI in a parliamentary context and built an automated pipeline for the current workflow.

Cheesecake Technologies LLC (02/2023- Present) — Founder and Software Engineer

• Designed and implemented serverless cloud infrastructure on AWS (Lambda, Cognito, RDS, S3) for regulated industry clients
• Built modular backend architecture with Node.js, Prisma, and PostgreSQL for structured, auditable data storage
• Led end-to-end development, deployment, and monitoring in a small agile team

DROM Solutions (10/2021-02/2023) — Working Student IT Project Manager

• Developed internal tooling with Microsoft Power Apps for SAP authorisation and GRC compliance processes
• Managed user and role administration in SAP systems with audit-proof documentation

STRAFFR GmbH (03/2020-09/2021) — Working Student Backend Developer (IoT Healthcare Startup)

• Built ML models for personalised fitness recommendations from IoT smart band sensor data stored in Firebase
• Optimised Firebase data reads and writes for real-time sensor data processing from wearable devices
• Evaluated and selected production ML tooling including Weka and Google AI Studio for model implementation

Shipd.ai by Datacurve, March 2026.

Emotion Recognition from Audio Data — TU Hamburg CNN-based classification model for emotion detection from audio signals focusing on on model quality and deployment in sensitive application contexts using Python and CNN → Speech Emotion Recognition Using Convolutional Neural Network and Its Use Case in Digital Healthcare My thesis was turned into a mental health project, called Bolo

RAG Hybrid Search Engine Semantic search engine with hybrid retrieval. Loads any PDF, chunks and embeds with Ollama (nomic-embed-text, 768-dim vectors), stores in Qdrant Cloud, searches with Vector + BM25 combined, ranks with Reciprocal Rank Fusion. Built as part of a 30-day AI engineering sprint. → QdrantRAG

HVV Ticket Fallback System Resilient digital ticketing backend with layered fallback mechanisms for device failure scenarios. Includes Architecture Decision Records and post-mortem documentation. → hvv-ticket-fallback

Languages

Backend

ML and AI

Cloud and Infrastructure

Databases and Vector Stores

Data Engineering