Current Focus
In the current period I have stayed active through maintenance, smaller technical work, portfolio development, and personal side projects while sharpening where I want to take my experience next. The common thread is becoming clearer: I want to build useful systems with more ownership, stronger product context, and a closer connection to real-world outcomes.
After years of working across platforms, dashboards, e-commerce flows, developer tooling, operational interfaces, and physical/digital prototypes, I am ready to bring that range into a product or engineering role. I am especially interested in internal tools, workflow platforms, civic or public-service-adjacent technology, health, housing, energy, infrastructure, connected products, and practical AI-assisted workflows.
The right next team would value thoughtful engineering, good design, clear communication, and long-term maintainability. I want to stay hands-on while helping shape products and services that make complex systems easier to understand, trust, and use.
Below are two projects in progress, with screenshots and a few notes on what I’m building and why.
Safepoint NEW
Application overview: the model proposes changes; a person reviews what would change and why.Promotion-release review: proposed changes, supporting evidence, and items needing attention.Rulebook prototype: exploring model-drafted rules and calculations against captured sheet data. - Stack:
- React
- Next.js
- TypeScript
- Vercel AI SDK
Safepoint is an independent product and interface project focused on reviewing AI-proposed changes before they reach operational systems. I designed and built the current promotion-release prototype around a clear review flow: what would change, which evidence supports it, and where a reviewer needs to make a decision.
The work combines a React and Next.js interface with TypeScript models, synthetic test data, and an evolving LLM and rule-checking workflow. The emphasis is on making complex decisions understandable while keeping model suggestions, deterministic checks, and human approval separate.
Street Reports NEW
Clustered map: nearby street issues grouped for browsing.
Report detail: photo, status, and location in a focused sheet.
Area summary: report counts and a non-model fallback insight.- Stack:
- Swift/SwiftUI
- MapKit
- Apple Foundation Models
Street Reports is a native iOS side project for exploring local civic issues using public FixMyStreet data. I designed and built a map-first SwiftUI experience with clustered reports, location search, and detail sheets that bring photos, status, and location into one focused view.
The project also explores on-device AI through Apple Foundation Models, structured output, and local tools, with a non-AI fallback when the model is unavailable. Networking, caching, and insight generation are kept separate, with tests covering data handling and validation.