Systems for verification, capability, and value under adversarial conditions
I build and originate the systems, platforms, and public properties that have to remain legible when trust degrades: protocols, ML infrastructure, knowledge platforms, long-form writing, and formal research. My work sits at the boundary between computation, institutions, culture, and lived capability.
- 7+ years across secure and decentralized systems
- Sub-50ms operational inference on constrained hardware
- Distributed ML in classified and air-gapped environments
- Peer-reviewed research and long-form public writing
What I Build
The work has several public faces, but the underlying question is the same: what still works when soft guarantees fail?
Protocols & Verification
Threat models, incentive design, proofs, privacy-preserving architecture, and systems built to remain credible when counterparties are strategic.
ML Infrastructure & Edge Systems
Distributed training, secure deployment, latency-sensitive inference, sensor fusion, and production workflows designed for actual operating conditions.
Knowledge Systems & Public Capability
Platforms that turn scattered information into usable skill — across water, shelter, power, repair, and community — through structured retrieval, versioned knowledge objects, and clear paths from need to action.
Writing, Money & Civilizational Systems
Long-form work on verification collapse, synthetic media, repression, sovereignty, money, coordination, and the search for durable value.
Quantitative Systems & Optimization
Mathematical optimization applied to financial construction — MILP solvers, payoff engineering, multi-dimensional scoring, and provably good structures over heuristic familiarity.
Music, Form & Research
Formal work in musical geometry, sonification, and structure — the research thread underneath how I think about pattern, invariants, and lawful motion.
Collective Intelligence, Agents & Retrieval
Multi-agent research platforms, enterprise context graphs, agentic retrieval planners, and knowledge systems where relevance is defined by causality and organizational state — not vector similarity alone.
Featured Properties
These are properties I founded and maintain — thesis, essays, research, capability infrastructure, and applied systems — each originated from scratch as part of the same architecture.
A book-length thesis on privacy, proofs, and verified compute as monetary primitives for a dense digital civilization.
Publication / Essays Eschatology ReportEssays and dossiers on AI, culture, institutional drift, symbolic overload, and the search for workable orientation.
Public Infrastructure / Knowledge Platform Capability CommonsA structured knowledge platform for practical public capability across water, food, shelter, power, repair, and community.
Research Program / Formal Mathematics GAMUTA formal research program on the geometry of musical possibility, from pitch content and cyclic order to symplectic structure.
Applied Systems / Commercial AlchemicalAIThe applied and commercial surface for selected architecture, infrastructure, and systems work.
Selected Work Preview
A few examples of the broader thesis expressed in shipped systems, operational constraints, and public infrastructure.
- Adversarial Storage & Incentives Built and led a ciphertext-only storage network where storage is proved over time rather than claimed once, paired with long-horizon incentive and reputation design.
- Real-Time Detection & Edge Inference Led hybrid DSP + ML systems for latency-sensitive classification on edge hardware, optimized around operational metrics rather than lab-only accuracy.
- Secure Distributed ML Infrastructure Built distributed fine-tuning and serving workflows inside air-gapped environments where every dependency, artifact, and operator workflow had to justify itself.
- Capability Commons Designed a Postgres-first knowledge platform for practical literacy across water, food, shelter, power, repair, and community-scale competence.
Some people arrive here through protocols. Others through writing, research, knowledge systems, or music. The underlying problem is the same: when persuasion scales faster than proof, what structures still let people verify reality, coordinate action, and preserve value?
Some of my work answers that in code. Some answers it in writing. Some answers it in formal research or public capability infrastructure. To me these are not separate careers. They are different surfaces of the same question.
Start with shipped systems. Then follow the thread into the writing, the research, and the longer-lived labs.