Projects

A structured approach to turning market signals into consistent directional views and scenarios. Designed to make assumptions explicit, track what changes them, and communicate outcomes clearly.

Integrated balance structure that connects production, operating rates, trade flows, and schedules into a coherent view of market tightness and risk.

Repeatable monitoring workflow that reduces manual reporting and helps focus attention on the few indicators that actually move the story.

Structured templates that keep outputs consistent across cycles and reduce friction when markets change quickly.

Built a multi-room HVAC control approach using weighted temperature averaging, priority overrides (e.g., nursery comfort), and sensor aggregation to balance comfort and energy efficiency. Implemented with a mix of YAML automations and Node-RED flows to keep the logic explicit, testable, and repeatable.

Integrated a gas meter signal into Home Assistant and built a dashboard to track consumption and make billing/usage trends easy to understand. Focused on reliability (signal capture), clean entities, and meaningful presentation.

A small homelab environment used to run services and experiment safely: virtual machines, resource planning, segmentation, and maintenance workflows.

Set up a local language model service on a mini PC and integrated it into personal tooling for experimentation. Focus areas include privacy, latency, reliability, and making AI useful in real workflows (not just demos).

Built a Discord bot to automate community workflows and add quality of life features. This project is a good example of iterative development: add features, observe behavior, refine.

A collection of small electronics and automation builds: sensors, control logic, and practical household upgrades. I enjoy the mix of constraints — power, reliability, and clean wiring — and making it all work end-to-end.