Shlomo Kashani · Deep-tech AI, quantum & strategic advisory
I bring the board-level judgement to choose the right AI bet and the engineering to build it, in one person. I build instruments to disprove my own hypotheses, and I trust a theory only once it survives a genuine attempt to kill it.
I take on a small number of chief executives and deep-tech teams as a Fractional Head of AI at a time — on the AI, quantum, and machine-learning decisions where getting it wrong is expensive. The fastest route to a serious conversation is a short, confidential note describing the decision in front of you and the timeline you are working to. I reply personally.
Where to start:
What I do
Beyond the advisory work, my first love is AI and quantum-physics research, and most of it ships as open source. Recent releases include QuantumNovelty, an agent framework that audits quantum-computing papers, alongside mlxQ, a quantum-simulation framework tuned for Apple Silicon, and QuantumLLMInstruct, a corpus of AI-generated quantum-computing problems and solutions.
Production large language models, retrieval, and agentic AI — designed, built, and shipped into real use.
Applied quantum computing and QML — research, prototyping, and simulation, including frameworks tuned for Apple Silicon.
Judgement on high-stakes AI decisions made under uncertainty and time pressure, for boards and chief executives.
Background
I study Defence and Strategic Studies and I am the practitioner who builds the systems — production LLM, retrieval, and agentic AI through to quantum computing — so I can hold the strategic frame and, in the same conversation, look under the hood of the specific claim in front of you.
The pace of AI capability is changing what institutions can build — and what they should refuse to build. The decisions are large, often irreversible, and made under time pressure, where moving too slowly is now as costly as moving recklessly.