I build practical software with AI, while developing the engineering judgment needed to use AI responsibly.
I'm a developer focused on AI-assisted development, useful automation, and products that solve real problems. I use AI to move faster across research, implementation, debugging, testing, and review—but I do not treat generated code as automatically correct.
My projects are designed to demonstrate more than technical output. They show how I approach unclear problems, challenge assumptions, protect against failure, and carry an idea through to documentation, testing, packaging, and public release.
- AI literacy: choosing where AI helps, where human judgment matters, and how to verify the result
- Engineering mindset: critical thinking, threat modelling, failure testing, redundancy, and attention to detail
- Product focus: making technically complex tools understandable and useful to real people
- Open to: software-development opportunities, collaboration, feedback, and thoughtful technical conversations
| Project | What it does | What it demonstrates |
|---|---|---|
| Carry | An open-source Windows app for safely syncing project folders and shared AI-agent memory between trusted computers. It protects work with conflict handling, recovery checkpoints, and careful failure testing. | Product development · Security thinking · Reliability · Tauri · Node.js · CI/CD |
| Shared Agent Memory | One shared, local memory and coordination layer for multiple AI coding agents. Claude Code, Codex, Cursor, Windsurf, Gemini CLI, Aider, and other MCP-capable tools can reuse context and coordinate file edits across sessions. | MCP · Agent orchestration · Developer tooling · Node.js |
| Sentinel | A safety-first Linux assistant that turns plain English into one proposed terminal command, filters destructive operations, and requires human approval before anything runs. | Python · LLM APIs · Guardrails · Human-in-the-loop design |
| Miko | A Windows voice assistant powered by Gemini Live that can work with the PC, Discord, the web, and local files. It also exposes tools that other agents can call. | Voice AI · Tool calling · FastAPI · Automation · Python |
| Beta Testing App | A Windows overlay for beta testers to manage test lists, capture OBS timestamps, and keep useful controls visible while testing another application. | Desktop UX · MVC architecture · OBS WebSocket · Supabase · Python |
I treat AI as a capable development partner, not autopilot. My usual workflow looks like this:
- Define the real problem and the behaviour the product needs to guarantee.
- Use AI to accelerate exploration and implementation while keeping control of architecture and trade-offs.
- Challenge the result with edge cases, security questions, interrupted workflows, corrupt inputs, and tests that try to prove it wrong.
- Add redundant protection where failure could damage user data or trust.
- Finish the product work through accessibility, documentation, packaging, release automation, and honest limitations.
This is the kind of AI literacy I want my portfolio to show: knowing how to get value from AI while remaining accountable for the outcome.
AI / LLMs:
LLM APIs · Tool calling · MCP · AI agents · AI-assisted development
CompTIA Tech+ Certification CompTIA View credential ↗ |
Google IT Support Coursera · Google View credential ↗ |
Linux Unhatched Cisco Networking Academy View credential ↗ |
IT Customer Support Basics Cisco Networking Academy View credential ↗ |
I'm actively building, learning, and looking for opportunities where practical AI literacy, product thinking, and careful software development are valued.
If you have feedback on one of my projects, a code-review observation, a question about my process, or a role that may be a good fit, I would be glad to hear from you on LinkedIn or by email.
⚡ This profile evolves with every project I build and every assumption I learn to test.