Independent researcher and software engineer building reproducible systems for governance, auditability, simulation, and AI-assisted reasoning.
I am most interested in work where rigorous models meet messy public problems: legal equality, institutional resilience, policy decisioning, knowledge systems, climate co-operation, and deterministic simulation.
| Project | What it shows | Status |
|---|---|---|
| Isonomia | Data, code, and papers modelling equality under law and governance resilience across historical systems | Active research |
| Ananke | Deterministic TypeScript simulation kernel with replay, stable API boundaries, conformance tests, and host-owned integration | Active engineering |
| Mnemosyne | Language learning from any text — FSRS scheduling, morphological analysis, 17 language plugins, self-hosted | Active product |
| Kalmanorix | Specialist embedding routing and reproducible retrieval evaluation toolkit | Active research |
| SENA | Deterministic policy-as-code engine for auditable Jira and ServiceNow approval workflows | Alpha prototype |
- Isonomia — a formal model of political freedom, legal equality, and governance resilience across roughly 2,600 years of historical systems.
- Tipping the climate equilibrium — tensor game theory and reinforcement-learning experiments for climate coalition discovery.
- Fixed points in time tensors — supporting code for work on constrained spacetime events and chronology protection.
- SENA — deterministic policy evaluation for enterprise approval workflows, with replayable audit evidence.
- Kalmanorix — specialist embedding routing and reproducible retrieval evaluation.
- Koios — exploratory ontology-grounded transformer architecture for interpretable reasoning.
- Veloclade — knowledge graph growth control using clade-style hierarchy.
- Ananke — deterministic simulation for imagined worlds, designed to sit inside a host-owned rendering, networking, and persistence stack.
- The Unfinished World — narrative game built around philosophical choice and world-state consequences.
- Mnemosyne — paste any text you care about — an article, a book chapter, song lyrics — and learn the language from it. Morphological analysis, per-word micro-lessons, and FSRS-scheduled review across 17 languages. Self-hosted.
- Nutri-Matic — micronutrient analysis: protein quality via DIAAS/PDCAAS, essential fats, fibre, vitamins, minerals.
- Atropos — LLM pruning and return-on-investment modelling.
- Code Green Code — energy-aware code auditing with CO₂ impact tracking.
Some repositories are working software, some are paper-support repositories, and some are research prototypes. I try to make that distinction explicit in each README.
The common thread is not a single framework or language. It is the attempt to make complex claims testable, replayable, and inspectable.
Dr Paul Schleifer
ORCID 0009-0004-7972-3566 · London, UK