Managing Partner and CTO building applied AI and autonomous delivery systems
I combine technology direction with hands-on Rust, Python and TypeScript engineering. As an inspiring technical leader, I build teams and strengthen performance, resilience, compliance and security. I lead the Terraphim AI privacy-preserving ecosystem and autonomous delivery infrastructure, alongside client architecture at Zestic AI. My experience includes enterprise architecture and the Synthetic Data Platform at Nationwide, industrial AI leadership at Klarian, startup team building at Shopitize and applied research at Cranfield. I am the author of Context Engineering: Designing AI That Understands, Remembers, Acts and Learns (2026).
A privacy-preserving AI ecosystem built in Rust and WebAssembly, connecting local-first knowledge graphs, shared agent context, editor workflows, command-line tools and MCP integration. Inspectable graph embeddings combine curated concepts, synonym mappings and graph co-occurrence. Terraphim private-cloud infrastructure uses a Rust control plane around Firecracker microVMs for lifecycle management, resource allocation, tenant routing, snapshots and operational telemetry.
Technical documentation site for the Rust engine and bindings behind Terraphim AI. Covers the knowledge-graph engine, deterministic Aho-Corasick matching and WebAssembly bindings, with install paths through cargo, Homebrew and curl. Documentation for the same engine rather than a separate product.
Cloud code forge built for AI agents, where they triage, coordinate and execute work. PageRank issue prioritisation, an MCP server for agent coordination and a Robot API for automation, running on isolated development environments. Hosted at terraphim-forge.com, with a public demo open.
Public catalogue for discovering, installing and publishing skills for AI coding agents, alongside the anonymous Skills MCP endpoint that serves signed and entitlement-aware content. The marketplace surface for the Terraphim Skills packages, with guides for each supported agent and editor.
I built and operate the AI Dark Factory, connecting prioritised issues and agent execution with separate review, verification and validation gates. Human owners retain control of acceptance and merge.
Rust platform of high-fidelity mock implementations of third-party APIs, for offline development, deterministic tests and cost-free experimentation. Each twin is a drop-in replacement compatible with the official SDK, with mirrored bearer-token auth, a webhook event bus, snapshot and restore for reproducible test states, and an endpoint-coverage dashboard. A shared twin-core crate carries error handling, pagination and clock abstraction across the services.
Machine Learning NLP pipeline for knowledge discovery using Redis ecosystem. Combines VR/AR interfaces with AI to fight information complexity and enable spatial information exploration.
Led team to build innovative distributed system resulting in a technology patent. Demonstrated systematic approach to innovation and ability to inspire teams to achieve breakthrough results.
I extended a Gitea-based development platform with scoped agent identity, signed audit events, a unified event stream and branch-based collaboration. A Rust bridge connects the platform to a kanban workflow for task coordination and review.
The signed skill packages and CLI for AI coding assistants, installable with one command. Disciplined engineering, Rust performance, DevOps and agent orchestration skills, cryptographically signed, compatible with Claude Code, Cursor, Codex and OpenCode. The catalogue that serves them is terraphim-skills.md.
AI-native time tracking via CLI. Login, create projects and tasks, start and stop timers, log time with natural language, generate reports, sync to cloud. Local-first with zero config. Rust core with Cloudflare Workers API, Tauri desktop app, and MCP server.
Designing AI That Understands, Remembers, Acts and Learns
Release candidate v1.0.0-rc1, updated 17 October 2026
A book on the discipline behind AI software that works. It argues that most AI project failures are context failures, not model failures, and sets out a practice: model the domain, ground answers in an explicit knowledge graph, budget context deliberately, give agents memory and learning that can be reviewed, and keep a named person accountable for what gets accepted. The argument runs from why AI software fails differently, through knowledge graphs and deterministic search, to evaluation, controlled action and enterprise governance.
Zestic AI | Jul 2023 - Present
I set technology direction and delivery priorities across Zestic AI’s client portfolio, combining hands-on engineering with reusable AI infrastructure and staged reviews. For a confidential digital-learning provider, I led pre-signature discovery and solution design, then architecture and engineering delivery. I designed governance-agent separation and specified deployment controls for technical approval, and delivered handover documentation covering rollback, cost guardrails, secrets and validation. Other work includes connected lending workflows for a confidential impact-finance lender.
Klarian | Nov 2023 - Jun 2025
Led AI and technology development for industrial pipeline monitoring and analytics, combining engineering models and machine learning to investigate pump efficiency and equipment anomalies. Co-authored published case studies on pump-performance analysis and orifice-plate anomaly detection.
Klarian | Dec 2022 - Dec 2023
InfinyOn | Dec 2022 - Dec 2023
Concurrent contract appointment.
Applied Knowledge Systems | Dec 2022 - Present
Concurrent self-employed appointment.
Nationwide Building Society Ltd | Jan 2019 - Dec 2022
Led the Synthetic Data Platform from inception to enterprise capability to support data-driven innovation and reduce reliance on live data. Developed privacy-preserving machine-learning and digital-twin approaches. Contributed to AI strategy and capability roadmaps covering ethics and accountable ownership; co-analysed cyber-physical systems for the Technology Radar and advised Credit Risk on AI feasibility and limitations.
Nationwide Building Society Ltd | Jan 2018 - Jan 2019
Established and ran a prototyping engineering lab, developed distributed-infrastructure prototypes and contributed to AI and blockchain strategy. Sponsored and mentored the team member whose work resulted in a distributed-data-storage patent.
Nationwide Building Society Ltd | Jan 2017 - Jan 2018
Shaped the enterprise DevOps blueprint, demonstrated architecture as code through a proof of concept and proposed a common business-value measure for change delivery.
Nationwide Building Society Ltd | Dec 2014 - Jan 2017
Led teams of 2–12 technical architects across strategic banking and regulatory-change programmes. Governed supplier designs and delivery quality, aligned architecture with enterprise strategy and challenged technical debt.
M2M Ecosphere | Jan 2019 - Jan 2021
Concurrent appointment with Nationwide Building Society Ltd. Established innovation processes, an R&D ecosystem, knowledge management and organisation design across Group Olvia. Deployed an AI prototype for the Group’s copper supply chain from the Democratic Republic of the Congo to China, addressing route optimisation and response to unexpected disruption.
Shopitize Ltd | Nov 2011 - Dec 2014
Served as the most senior technology professional, aligning business strategy with product development. Led distributed engineering and design teams and built an in-house engineering team. Owned architecture across automated receipt processing, OCR, product search, mobile APIs, backend services and cloud infrastructure. Mentored data scientists and was a named co-inventor on patent application WO2013057491.
Cranfield University | May 2007 - Jan 2012
Developed wavelet-packet methods for mobile spectrum sharing and implemented the waveform on signal-generation and analysis hardware. Researched communications for in-vehicle sensor networks in an aerospace health-monitoring programme supported by a consortium led by Rolls-Royce and Boeing.
Microsharp Corporation Limited | Feb 2003 - May 2007
Design, development and testing of IT and phone infrastructure, created fully automated and redundant IT infrastructure.
System Administrator magazine (samag.ru) | Feb 2002 - Feb 2003
Responsible for developing and controlling the publishing process. Management of a small team (5-8 employees) ensuring timely delivery to the printer.
Interested in collaborating on the convergence of AI, ethics, and systems engineering? Let's discuss how we can build the future of intelligent systems together.
Areas of Interest:
Natural Language Processing, Knowledge Graphs, VR/AR Interfaces, Data Privacy, Synthetic Data, Distributed Systems, AI Ethics, Healthcare AI, Biotech Applications, Enterprise Architecture
"Meta needs Brain -- Cortex: combining AI, ethics, systems and enterprise engineering to solve real-world complexity."