Deploy & compare local models
GGUF / llama.cpp / Ollama-style runtimes, quantisation choices, context, VRAM, latency and practical fit across NVIDIA and AMD hardware.
I came to AI from banking, compliance, digital identity and transactions. I founded Arival Bank. My work asks how institutions establish who may act, on whose behalf, with which evidence — and how to keep that meaning when software acts for them.
The difference between a system accepting an instruction and an institution being entitled to act on it.
NVIDIA Claw Agent Challenge: London, 2026. A fictional-business payments experiment; recognition of the demo, not endorsement of the book.
Project and evidence ↗Each volume hands the same anxiety to a different actor — fear to the self, power to the industry, ambition to a licensed fintech.
Futuroshock metabolises the fear of AI that arrived before any audit could.
→ OPEN THE PREQUEL BThe NVIDIA Innovator's Dilemma audits the empire — and maps what could still be built.
→ OPEN THE MAIN CEurope's First Trillion-Dollar Startup? — from financial operations to the authority behind AI actions.
→ OPEN THE SEQUELTechnical ability measured not by memorised syntax, but by framing the problem, specifying constraints, directing the build, inspecting the evidence and getting to a reproducible artefact.
GGUF / llama.cpp / Ollama-style runtimes, quantisation choices, context, VRAM, latency and practical fit across NVIDIA and AMD hardware.
Use cloud reasoning as the planning/review layer while Linux-hosted CLI agents, local models and services do the execution.
Baselines, matched prompts, controlled conditions, synthetic cases, regression checks, raw evidence and explicit failure criteria.
Deterministic validation → First-Order Logic / Prolog → institutional policy / OPA, with AI kept outside the final authority path.
Local STT/TTS, image generation/editing, research pipelines and specialist model roles rather than forcing one LLM to do everything.
APIs, data models, onboarding logic, web/mobile surfaces, test suites and public technical artefacts — without pretending every layer was hand-typed.
A document, an executable test, a counterexample. The point is not the number of computers on the desk.
Infrastructure, evaluation, identity systems, formal reasoning and AI-assisted product builds.
Local-model and multi-node experiments exploring role separation, heterogeneous hardware, and cloud-assisted planning with local execution. Individual setups are dated in the linked research.
AI-assisted self-sovereign identity reference system with web + mobile surfaces, PostgreSQL, OpenAPI 3.1, generated clients and a branching KYC/KYB onboarding engine.
Experiments around whether natural-language model outputs can be translated into explicit, inspectable logic and then constrained by deterministic or institutional policy layers.
A set of experiments asking a harder question than “can an agent act?”: under what evidence, mandate, scope and revocation state should an agent be allowed to act?
Local speech recognition and TTS in local experiments; ComfyUI image generation and editing in the local lab; specialist workloads kept separate from the main LLM path.
The domain behind much of the technical work: reusable identity, compliance-grade architecture, proof-of-personhood, government wallets, KYC/KYB and AI-era institutional trust.
The useful skill is turning a hard problem into an executable experiment, using specialist AI as leverage, and refusing to accept a green screen without evidence.
| Layer | Toolbox used in the experiments | What it is for |
|---|---|---|
| Local LLM | Qwen · Nemotron · Mistral · Gemma-class experiments · abliterated sandbox models | Inference, extraction, synthesis, coding, adversarial comparisons. |
| Runtime | llama.cpp · Ollama · LM Studio · FreeToken experiments | Local serving, GPU offload, quantised inference, heterogeneous hardware tests. |
| Agents | ChatGPT / Claude / Codex / Claude CLI · desktop/CLI hand-off patterns | Plan → execute → inspect → revise loops across cloud and local environments. |
| Formal | SWI-Prolog · First-Order Logic · OPA / Rego · Z3 model checks | Make assumptions explicit; keep policy and AI output separable. |
| Research | Local Deep Research · SearXNG-style search layer · local LLM synthesis | Evidence-led research without treating the base model as a current-information oracle. |
| Media | ComfyUI · FLUX-class image models · Whisper · local TTS | Local image generation/editing and speech workflows. |
| Product build | OpenAPI · React / Vite · React Native / Expo · PostgreSQL · Docker Compose | AI-assisted reference systems and prototypes with inspectable interfaces and tests. |
| Evidence | Git · GitHub Actions · hashes · manifests · raw outputs · regression suites | Turn a demo into something another person can inspect and rerun. |
Banking, fintech, compliance and digital identity taught him to care about who is allowed to do what, on whose evidence, under which rules.
Not a career change — the same profession.
Only the subject of trust changed: first the institution, now the model.
That is why the technical experiments keep converging on authority, evidence, provenance, identity, local control and regulated decision-making.
Not trying to become a better programmer than programmers — but very good at turning difficult institutional problems into systems that can be built, attacked, measured and inspected.
Most useful where product, AI, regulation, identity, infrastructure and institutional reality collide — and where a slide deck is not enough. The preferred output is something you can run.