6 public / 0 controlled
/Portfolio / AI
Projects that show practical AI use: bounded assistance, traceable decisions, auditability, validation, and clear limits on automation.
Private owner-only and internal tools are deliberately excluded from these public portfolio views. Protected entries do not expose private data on the portfolio page.
/public-apps
Open app pages, policies, store links, changelogs, and public case files.
Dialled MTB is a maintenance and setup app for mountain bikers, combining passive Strava ride sync, structured service tracking, setup calculations, and contextual AI support.
Uses contextual support to help riders make better setup choices and maintain their bikes with more confidence.
Sprocket is a plain-language phone helper for everyday support: ask by voice or text, set reminders, remember details, get step-by-step phone help, and make sense of confusing letters, emails, or texts.
Consumer AI-support pattern for calm phone help across voice and text assistance, reminders, memory, step-by-step tech guidance, read-aloud replies, and plain-English explanations.
RehabPath is a structured recovery companion for people working through injury rehab. It provides clear, phase-based programmes, step-by-step exercise guidance, and visible progress tracking—all stored privately on your device. No accounts, no tracking, no complexity. Just a calm, focused tool to help you stay on course and see your improvement.
Uses AI to ingest physio plans and produce clinician-friendly summaries.
/public-systems
Sanitised product concepts and frameworks that can be shared without internal access.
A closed-loop product system for discovery, backlog structure, testing, documentation, release context, and Ruby product intelligence.
Shows how product, delivery, testing, and docs can work as one traceable operating system.
AI system audit-trail infrastructure for provenance, observability, lineage, consent evidence, and compliance workflows.
Built around bounded accountability: what rules applied, who approved, what was believed, and what can be proven.
System of Record for Reasoning: a documented reasoning layer for constraints, decisions, findings, and safe AI-assisted delivery.
Useful as a personal brand signal because it demonstrates how AI work is governed, not just generated.