Co-founder & CTO at RFI-IRFOS, an interdisciplinary research facility in Graz, Austria.

I build AI systems from the substrate upward: models, inference infrastructure, agentic systems, and the tooling around them.

My work spans ternary AI, sparse inference, agent architectures, cybersecurity, and human-AI systems.

• Ternary intelligence — balanced {-1, 0, +1} neural networks, ternary model architectures, sparse inference, and the systems required to run them.
• AI infrastructure — inference runtimes, model tooling, retrieval systems, developer infrastructure, and research software.
• Agentic systems — autonomous systems for research, engineering, security, and intelligence workflows.
• Human-AI systems — investigating interaction, behaviour, emergence, and the boundary between human and machine decision-making.
• Cybersecurity — root-level investigation of software, infrastructure, and autonomous systems.

I work primarily across Rust, Python, C/C++, and TypeScript, and tend to take things from the research paper all the way down to the implementation.

Research and engineering infrastructure for ternary AI, including balanced ternary computation, model architectures, sparse execution, and evaluation.

A model-agnostic AI terminal environment built for working directly with local and remote models, providers, tools, and agentic workflows.

Vector and retrieval infrastructure built as part of our broader research into machine-readable knowledge and AI systems.

An experimental observatory for studying human-AI interaction, behavioural dynamics, and emergent system behaviour.

Language tooling for Ternlang, including parsing infrastructure and the foundations for a programming environment around ternary computation.

The standard library and runtime foundations for the Ternlang ecosystem.

Ternary computation · sparse inference · AI systems · agent architectures · cybersecurity · human-AI interaction · observational science · macro systems

fable was here — Claude Fable 5