AI Product · Automation · Agentic Systems
I build AI systems that do real work.
I turn business problems into working AI-enabled systems. As the founder of iTradeAIMS, I build and operate agentic workflows, automation, commercial platforms, licensing, customer operations and applied financial-ML systems - with a specification-driven, test-heavy approach to delivery.
What I build with
Tools matter, but the differentiator is the system around them: specifications, context, permissions, tests, review and deployment gates.
Builder first. Technology is the leverage.
I have spent my working life building and operating things. Earlier in my career that meant physical businesses - marble processing, flour milling, cold storage and warehousing. Later it became trading systems, digital products and iTradeAIMS. AI has expanded the scale of what I can build and the speed at which I can move from an idea to a working system.
Today I design AI-native product and engineering workflows: specifications, agent orchestration, MCP tools, reusable skills, test-driven implementation, code review, CI/CD gates and controlled production access. I use Codex and Claude for serious engineering work, Gemini for selected automation and content workflows, and deterministic code wherever deterministic behaviour is the better choice.
My strongest domain is FinTech and algorithmic trading, where I combine long trading experience with MQL4/5, Pine Script, Python, PyTorch, ONNX and web technologies. The same approach applies outside trading: identify the business bottleneck, design the system, automate the repetitive work and keep humans in the loop where judgement matters.
I am open to UK-remote roles in AI product engineering, automation, agentic systems, solutions engineering and applied AI. I am particularly useful where a company needs someone who can bridge business context, product decisions and hands-on technical delivery.
Flagship Case Studies
Production systems, operating workflows and applied AI work. The emphasis is on the problem, the controls around the system and what changed in practice.
Problem
Rebuild a reacquired FinTech business and remove routine operational work.
What I built / contributed
Rebuilt iTradeAIMS into an API-driven commercial platform spanning payments, entitlements, licensing, customer access, email and TradingView fulfilment.
Outcome / scope
Routine product fulfilment now runs effectively without manual handling after a customer purchase.
Problem
Protect commercial MQL software while eliminating manual licence fulfilment.
What I built / contributed
Laravel/API-driven licensing system that generates software entitlements after purchase and surfaces them in the iTradeAIMS member portal.
Outcome / scope
Licensing is integrated into the purchase-to-access workflow rather than handled as a separate manual task.
Problem
Make powerful coding agents reliable across a complex multi-repository production estate.
What I built / contributed
Private skills and plugin stack spanning Codex, Claude Code, Cursor and Antigravity, with session bootstraps, TDD, code review, domain skills, guardrails and delivery gates.
Outcome / scope
Reusable engineering discipline and domain context are carried across agent harnesses instead of being recreated prompt by prompt.
Problem
Test whether ML improves an existing trading method without being fooled by leakage or overfitting.
What I built / contributed
Directed and operated connected Edge Hunt, Train Model and AIMS Quant workflows for statistical evidence, point-in-time features, PyTorch training, adversarial leakage tests, ONNX export and MT5 integration.
Outcome / scope
A research-to-deployment path links validation controls to model artefacts and native MetaTrader inference.
Problem
Automate repetitive business work while retaining deterministic controls and human escalation.
What I built / contributed
Built governed content workflows and previously built a Discord support automation architecture combining deterministic intake with AI reasoning, retrieval, severity routing and escalation.
Outcome / scope
The content factory remains an active workflow; the original support implementation was later superseded and is retained as architectural evidence.
Problem
Bridge TradingView strategy signals into MT4/MT5 execution workflows.
What I built / contributed
Originated the product concept and serve as the trading/FinTech domain expert, defining user workflows, requirements and product direction.
Outcome / scope
My role is product and domain leadership; the startup engineering team owns implementation.
Let's build something useful
Looking for someone who can turn an AI idea into a working system?
I am open to UK-remote permanent roles, full-time or part-time, across AI product engineering, automation, agentic systems, solutions engineering and applied AI. Occasional travel is fine. I am comfortable working from business problem definition through architecture, implementation, testing and production delivery.
I can help with
- Agentic workflow design and orchestration
- Business-process and operational automation
- AI-enabled product engineering
- APIs, integrations and production workflows
My working style
- Specification-driven and outcome focused
- TDD, code review and explicit guardrails
- Human approval for high-impact actions
- Pragmatic: deterministic code where it is better than AI
FinTech is my deepest domain, but I am not limited to it. The common thread is solving operational and product problems with AI, automation and software.