Applied AI · End-to-end product engineering

David Horvath

I build AI-powered products end to end — from training the models to shipping the software and hardware they run on.

Currently: sole engineer at a bio-3D-printing startup in Vienna.

Selected work

Things that shipped

Three systems, written up the way an engineer would explain them: the constraints, the decisions, and what runs in production today.

A bioprinter nozzle extruding material into a culture vessel

Bio3D printer software

The complete software stack for a shipping bioprinter: frontend to firmware boundary, four devices, offline updates, one engineer.

  • React/TypeScript
  • C#
  • Jetson
  • CI/CD
HírSpektrum wordmark over the tagline “Nem hír. Narratíva.”

hirspektrum.hu

A self-hosted media analysis platform that clusters the Hungarian news feed and generates LLM-powered comparative analysis.

  • RabbitMQ
  • Neo4j
  • LLM/RAG
  • Clustering
A neural network visualisation classifying a handwritten digit

Vision to LLM, in production

Five years of applied AI at scale: real-time video pipelines on AWS, custom YOLO models, RAG systems, and LLM agents for financial analysis.

  • YOLO
  • PyTorch
  • AWS
  • Palantir Foundry

The Lab

Where computer vision goes to dance

Interactive installations, live audio-reactive VJ sets, and real-time experiments. Shown at Ars Electronica and the Zsolnay Light Festival.

Enter the Lab

About

David Horvath

Five years building software, four of them building AI-powered products end to end — computer vision at Verizon, LLM and agent systems at Swiss Re, and now the entire stack of a shipping hardware product.

I like the parts other people hand off: the device boundary, the release process, the architecture decision nobody wants to own.