Projects
Most of these started because a workflow felt clumsy, repetitive, or harder than it needed to be. I like building tools that create more room for better judgment.
The automated system behind my own multi-platform publishing workflow, from raw recording to review-ready posts.
Why it exists
I needed a repeatable way to move from one recording to multiple platform outputs without rebuilding the process every week.
Current focus
Keeping approvals light, outputs consistent, and the whole system flexible enough to evolve with the brand.
A browser editor for turning long recordings into short, captioned clips without bouncing between five different tools.
Why it exists
I wanted a faster way to shape short-form video from transcripts, timing, and captions in one workflow.
Current focus
Tightening the timeline feel, transcript editing, and export polish so the app feels dependable instead of just clever.
A long-term home for projects, essays, music notes, and the context that would get flattened anywhere else.
Why it exists
Social platforms are useful, but they are not a stable home for the full picture.
Current focus
Making the structure stronger, giving writing room to grow, and keeping the voice honest.
A desktop app for surfacing the strongest moments inside longer videos before the review stage gets messy.
Why it exists
Manual clip hunting takes too long when the real job is deciding what is worth keeping.
Current focus
Improving moment scoring, cleaning the review flow, and making the picks easier to trust.
Computer-vision-assisted thumbnail selection from raw video frames.
Why it exists
The best frame is usually buried in too much footage and found too late.
Current focus
Balancing technical scoring with actual visual judgment so the recommendations feel useful instead of robotic.
A research tool for spotting themes and conversations before they become obvious.
Why it exists
I wanted a better way to notice what is gaining traction without scrolling aimlessly or guessing off vibes alone.
Current focus
Testing sources, filtering noise, and deciding which signals actually deserve follow-up.