Verified work · Current focus · Ecosystem · Standards · Research output · Collaboration
Independent researcher and systems engineer at the Anulum Institute in Switzerland.
I build evidence-governed infrastructure for AI systems, multi-agent engineering, scientific computing, neuromorphic hardware, quantum simulation, and control. The work connects mathematical models to reproducible software, native acceleration, formal models, and executable hardware paths.
Claims are only as good as the measurements, artefacts, or verification that support them.
| AI assurance Grounding, contradiction detection, action review, audit evidence |
Agent infrastructure Coordination, claims, durable messaging, memory, fleet control |
Scientific systems Plasma physics, oscillators, quantum workloads, numerical validation |
Compute to hardware Rust acceleration, FPGA RTL, WebGPU, formal verification |
Primary implementation
Extended scientific, formal, hardware, and operations stack
Scientific, native, and formal work
Hardware, web, and operations
The portfolio also contains maintained protobuf/gRPC contracts, Python-Rust bridges built with PyO3 and Maturin, WebAssembly targets, native SIMD paths, scientific notebooks, and multi-language API documentation.
| Goal | Entry point |
|---|---|
| Coordinate parallel coding agents without worktree collisions | Synapse Channel and its documentation |
| Guard LLM claims and factual consistency | Director-AI and its documentation |
| Audit repositories and prepare evidence-bound remediation | Rigor Foundry and its documentation |
| Build auditable memory for agents and knowledge systems | Remanentia |
| Explore neuromorphic and stochastic computing | SC-NeuroCore and its documentation |
| Work with plasma models and control-grade runtimes | SCPN Fusion Core and SCPN Control |
| Reproduce coupled-oscillator quantum experiments | SCPN Quantum Control |
These links lead to inspectable artefacts, not summary claims.
| Multi-agent coordination Claims, durable messaging, federation, and threat-bounded execution. Validation · Coordination specification · Threat model |
LLM response assurance NLI/RAG grounding, claim review, native acceleration, and declared capability boundaries. Validation · Public benchmarks · Capability matrix |
| Neuromorphic compute to RTL Python models, Rust SIMD paths, compiler surfaces, and hardware evidence. Validation · Synthesis results · Traceability matrix |
Plasma modelling and validation Solver evidence, real-data paths, uncertainty boundaries, and reproducible artefacts. Validation · DIII-D validation record · Software DOI |
| Quantum experiment governance Preregistration, hardware result packs, raw counts, and explicit non-advantage boundaries. Preregistration · Result-pack contract · Software DOI |
Research honesty in phase systems Matched-false-alarm evaluation, negative results, and bounded transfer claims. Submission index · Negative-result preprint · Grid regime-map preprint |
Portfolio state verified 2026-09-03.
| Reactor portfolio Consolidating shared kernels and governed device truth across 25 public reactor repositories. |
Agentic assurance Joining coordination, memory, response assurance, action review, and repository evidence without collapsing ownership boundaries. |
Research to hardware Carrying scientific models through native acceleration, formal checks, RTL, hardware runs, and inspectable result packs. |
| Period | Publicly supported milestone |
|---|---|
| 1996 onward | Self-published concept-development horizon across the broader research programme |
| 1998 | Founder role at ANULUM CH&LI begins in the self-entered public ORCID record |
| 2018 | Public GitHub account established |
| 2025 | Public SCPN previews, framework indexes, and technical reports deposited on Zenodo |
| 2026 | Public software and research portfolio expands across AI assurance, agent infrastructure, neuromorphic computing, plasma control, and quantum simulation |
Timeline entries distinguish registry facts from self-published chronology. They do not imply academic affiliation, external validation, funding, or awards.
| 01 · Inspect Rigor Foundry identifies broken, unproven, or unsafe-to-claim surfaces. |
02 · Guard Director-AI evaluates model output before it is trusted. |
03 · Coordinate Synapse Channel manages claims, mailboxes, plans, and receipts. |
04 · Compute SC-NeuroCore and SCPN execute scientific, physical, and hardware-facing work. |
Research, validation, and product readiness remain separate. Active development is not a readiness claim.
The map contains 39 portfolio repositories: 33 public repositories and six private product surfaces. HushLine is a standalone public utility outside the five portfolios. Connections represent contract, integration, evidence, and audit flow. They do not merge ownership or imply scientific validation, operational readiness, or actuation authority. Counts verified 2026-09-03.
Access key: PUBLIC · PUBLIC / ARCHITECTURE-ONLY · PRIVATE ·
PRIVATE / PROPRIETARY
01 · SCPN Reactor Systems 25 public repositories
Device-family physics, shared numerical kernels, reactor models, geometry, and configuration ownership. Repository presence does not by itself establish validated physics or machine readiness.
Thematic index: shared foundations (3) · magnetic confinement (10) · inertial confinement (3) · MIF and pulsed systems (4) · alternative concepts (5)
| Repository | Scope | Access |
|---|---|---|
| SCPN Reactor Kernels | Shared deterministic physics, geometry, and numerical kernels for the reactor portfolio | PUBLIC |
| SCPN Fusion Core | Tokamak physics, solvers, validation campaigns, transport, and control research | PUBLIC |
| SCPN MIF Core | Pulsed FRC/MIF kinematics, deterministic trigger logic, FPGA RTL, and formal timing evidence | PUBLIC |
| SCPN Beam Target Core | Fixed-target and colliding-beam fusion device truth | PUBLIC |
| SCPN Dense Plasma Focus Core | Coaxial dense-plasma-focus devices, diagnostics, and level-0 physics | PUBLIC |
| SCPN FRC Core | Field-reversed-configuration fusion systems | PUBLIC |
| SCPN Fusion-Fission Hybrid Core | Fusion neutron sources coupled to explicitly subcritical fission blankets | PUBLIC |
| SCPN ICF Beam Core | Ion- and pulsed-electron-beam inertial confinement fusion | PUBLIC |
| SCPN ICF Impact Core | Projectile- and impact-driven inertial confinement fusion | PUBLIC |
| SCPN ICF Laser Core | Direct-drive, indirect-drive, and staged laser inertial confinement fusion | PUBLIC |
| SCPN IEC Core | Gridded and Polywell-style inertial electrostatic confinement | PUBLIC |
| SCPN Lattice Fusion Core | Governed boundary for externally driven lattice-confinement fusion research | PUBLIC / ARCHITECTURE-ONLY |
| SCPN Levitated Dipole Core | Levitated-dipole confinement systems | PUBLIC |
| SCPN Magnetic Cusp Core | Purely magnetic cusp-confinement systems | PUBLIC |
| SCPN MIF Liner Core | Mechanical- and liquid-liner magnetised-target fusion | PUBLIC |
| SCPN MIF MagLIF Core | Premagnetised, laser-preheated, pulsed-power-driven MagLIF systems | PUBLIC |
| SCPN MIF Plasma Jet Core | Converging plasma-jet-liner magneto-inertial fusion systems | PUBLIC |
| SCPN Mirror Core | Simple, tandem, and gas-dynamic magnetic mirrors | PUBLIC |
| SCPN Muon Fusion Core | Governed boundary for muon-catalysed fusion research | PUBLIC / ARCHITECTURE-ONLY |
| SCPN RFP Core | Reversed-field-pinch fusion systems | PUBLIC |
| SCPN Spheromak Core | Self-organised spheromak compact toroids | PUBLIC |
| SCPN Stellarator Core | Stellarator, heliotron, and torsatron systems | PUBLIC |
| SCPN Theta Pinch Core | Theta-pinch devices, diagnostic contracts, and cited level-0 physics | PUBLIC |
| SCPN Tokamak Core | Configuration and diagnostic-plan truth for conventional and spherical tokamaks | PUBLIC |
| SCPN Z-Pinch Core | Classical and sheared-flow Z-pinches, level-0 physics, and deterministic geometry | PUBLIC |
02 · SCPN Systems Integration and Control 4 repositories
| Repository | Scope | Access |
|---|---|---|
| SCPN Control | Neuro-symbolic controllers, runtime admission, replay, audit, and software action boundaries | PUBLIC |
| SCPN Phase Orchestrator | Coupled-rhythm analysis, sealed synchronisation evidence, and review-only control proposals | PUBLIC |
| SCPN Studio | Federating Hub for scientific studios, claim boundaries, portfolio visibility, and gated execution | PRIVATE |
| SCPN Studio Platform | Domain-neutral SDK for evidence bundles, capability manifests, jobs, identity, and portfolio validation | PRIVATE / OPEN-CORE |
03 · Agentic Coordination, Assurance and Continuity 8 repositories
| Repository | Scope | Access |
|---|---|---|
| Director-AI | LLM hallucination guardrail with NLI/RAG grounding, sealed evidence, and optional streaming contradiction checks | PUBLIC / OPEN-CORE |
| Director Class AI | Pre-dispatch review and evidence controls for high-impact autonomous-agent actions | PRIVATE / BUSL-1.1 |
| Director AI Cloud | Managed multi-tenant accounts, API keys, metering, quotas, and hosted-service gating | PRIVATE / BUSL-1.1 |
| Rigor Foundry | Evidence-bound inventory, audit candidates, review binding, and remediation planning | PUBLIC |
| Remanentia | Auditable AI memory with hybrid retrieval, graphs, consolidation, CLI, MCP, and API surfaces | PUBLIC |
| Remanentia Portal | Customer identity, entitlement, and commercial control plane without customer-workload execution | PRIVATE |
| Synapse Channel | Local-first agent coordination with plans, claims, durable messaging, audit, and protocol adapters | PUBLIC |
| Synapse Channel Fleet | Licensed multi-machine federation, trust management, offline licence admission, and cross-hub operations | PRIVATE / PROPRIETARY |
04 · SC Neuromorphic Computing Systems 1 public repository
| Repository | Scope | Access |
|---|---|---|
| SC-NeuroCore | Stochastic and spiking neural systems with Python APIs, Rust acceleration, HDC/VSA, and RTL-generation workflows | PUBLIC |
05 · SCPN Quantum Computing Systems 1 public repository
| Repository | Scope | Access |
|---|---|---|
| SCPN Quantum Control | Coupled-oscillator quantum experiments, simulators, optimisation, hardware runs, and hash-bound result packs | PUBLIC |
Standalone utility 1 public repository
| Repository | Scope | Access |
|---|---|---|
| HushLine | Deterministic command wrapper that filters, bounds, and optionally redacts stdout and stderr | PUBLIC |
Practices are applied according to each repository's risk and scope. Not every repository runs every tool.
| Quality dimension | Practices used across the portfolio |
|---|---|
| Correctness | Deterministic pytest and Cargo suites, branch-aware coverage gates, parity tests, regression fixtures, and explicit negative cases |
| Static quality | Ruff formatting and linting, strict mypy where declared, Cargo fmt, Clippy with warnings denied, and API-contract checks |
| Reproducibility | Hash-pinned dependency locks, preregistered protocols, raw result packs, content digests, benchmark metadata, and replayable audit records |
| Security | Bandit, CodeQL and scorecards where enabled, threat models, least-authority execution, secret boundaries, and dependency review |
| Supply chain | SPDX headers, REUSE 3.x checks, SBOM generation where applicable, pinned CI actions, signed or digest-bound evidence, and release manifests |
| Polyglot verification | Python/Rust parity, PyO3 and Maturin bridges, Go and Julia tests, Lean builds, WebAssembly targets, and RTL/formal checks where relevant |
| Documentation | Warning-fatal or strict MkDocs/Sphinx builds, generated API references, architecture decisions, validation records, and explicit non-claims |
| Delivery | Repository-local preflight gates, CI workflows, PyPI packages, wheels and source distributions, containers, and benchmark harnesses |
The verified PyPI profile currently contains 19 published projects. They include public Python packages, Rust-accelerated engines, domain kernels, and command-line tools.
Published package index 19 projects
Agent and assurance systems: synapse-channel, director-ai, director-ai-lite, rigor-foundry, remanentia, backfire-kernel, and hushline.
SCPN systems: scpn-phase-orchestrator, spo-kernel, scpn-control, scpn-fusion, scpn-fusion-rs, scpn-mif-core, scpn-quantum-control, scpn-quantum-engine, oscillatools, and scpn-studio-platform.
Neuromorphic systems: sc-neurocore and sc-neurocore-engine.
| Surface | Verified route |
|---|---|
| Complete research index | Publications, preprints, software archives, and evidence boundaries |
| Curriculum vitae | One-page PDF · Markdown source · JSON Resume |
| Research identity | ORCID 0009-0009-3560-0851 |
| Software publication | 19 projects on PyPI |
| Quantum-control software | Zenodo DOI 10.5281/zenodo.18821929 |
| Fusion software | Zenodo DOI 10.5281/zenodo.18820864 |
| Phase-system preprints | Matched false-alarm study and grid regime-map study |
| HushLine software | Zenodo DOI 10.5281/zenodo.20775432 |
| Project | Role | Maturity |
|---|---|---|
| Synapse Channel | Control plane for coding-agent fleets with claims, roles, durable mailboxes, receipts, audit, and federation | Usable now: functional core, active development |
| Rigor Foundry | Evidence-bound repository auditing and remediation planning | Usable now: active hardening |
| Director-AI | Real-time LLM guardrails with NLI/RAG fact-checking and optional claim-level streaming halt | Research active: functional system under validation |
| SC-NeuroCore | Stochastic and neuromorphic framework spanning Python, Rust SIMD, Verilog, and HDC/VSA | Research active: platform under continuous development |
| SCPN Quantum Control | Evidence-governed quantum simulation of coupled-oscillator synchronisation | Experimental: preregistered research programme |
Dynamic badges report registry and workflow state. They are operational signals, not scientific-quality scores.
| Project | Release | Main CI |
|---|---|---|
| Synapse Channel |  |
 |
| Rigor Foundry |  |
 |
| Director-AI |  |
 |
| SC-NeuroCore |  |
 |
| SCPN Quantum Control |  |
 |
| Label | Meaning |
|---|---|
| Usable now | Installable, documented, and CI-backed; still evolving |
| Research active | Real code and ongoing science; not a stability promise |
| Experimental | Exploratory; interfaces and claims are not fixed |
| Evidence-bound | Public claims are tied to measurements or artefacts |
Negative and null results are published when they are real. Public claims stay tied to measurements, preregistered protocols, raw packs, source-bound reviews, or executable verification.
Example: SCPN Quantum Control publishes preregistered protocols and hash-bound result packs rather than converting experimental activity into a readiness or advantage claim.
- Evidence before claims.
- Reproducible artefacts before presentation.
- Clear boundaries between research, validation, and product readiness.
- Cross-language implementations where performance or hardware integration justifies them.
- Fail closed when provenance, authority, or evidence is incomplete.
- Negative results and failure records remain part of the research output.
| Model | Typical boundary |
|---|---|
| Apache-2.0 | Permissive public cores such as Director-AI and Rigor Foundry |
| AGPL-3.0-or-later | Public network-facing and research systems with source-sharing obligations |
| Open core | Public core with separately licensed advanced or managed product surfaces |
| BUSL-1.1 | Selected private commercial systems with a declared future change licence |
| Commercial licence | Alternative terms for organisations that cannot use the public licence |
| Mode | Scope |
|---|---|
| Research collaboration | Reproducible studies in AI assurance, neuromorphic systems, quantum simulation, plasma physics, and control |
| Technical collaboration | Architecture review, validation design, formal or hardware paths, and evidence-bound software engineering |
| Commercial licensing | Dual-licensed and managed product surfaces through the Anulum licensing route |
| Open-work sponsorship | CI, compute, hardware and quantum experiment time, and public documentation through GitHub Sponsors |
I welcome technically grounded collaboration in reliable AI infrastructure, multi-agent systems, neuromorphic computing, scientific software, formal verification, quantum simulation, plasma physics, and control.
Engagements are taken selectively. The public GitHub profile currently marks the account as hireable; this is not a guarantee of immediate capacity.
A useful first message includes the problem, constraints, relevant prior art, and what evidence would count as success. Contact me through protoscience@anulum.li or anulum.li.
I respond to technical proposals. I do not take on ungrounded hype work, demo-only science theatre, or claims that cannot be checked.
For sustained open work, GitHub Sponsors funds CI runners, compute, hardware and quantum experiment time, and public documentation rather than marketing.
Transparency: These repositories span research software, developer tools, private systems, and product candidates. Active development does not imply production readiness or scientific validation unless a project provides explicit evidence.
I AM THAT
