I'm not a traditional developer.

I'm a writer, researcher, designer, archivist, obsessive documenter, and enthusiastic maker who discovered that not knowing how to build something is no longer a particularly compelling reason not to try.

I work heavily with AI.

I learn technologies when a project gives me a reason to learn them.

The project comes first.

The stack follows.

I don't really have one category.

• Sometimes I identify a problem.

• Sometimes I discover a dataset.

• Sometimes I become annoyed with existing software.

These are all equally dangerous.

An experiment in treating documents as modular systems rather than one long stream of markup.

Part language design, part publishing experiment, part attempt to answer:

What would markup look like if I designed it around the way I actually think?

Tools for turning years of accumulated email into something you can actually explore.

Search it. Classify it. Find forgotten relationships, projects, conversations, patterns, and pieces of your own history buried inside an inbox.

Your email archive is a dataset.

Might as well interrogate it.

Interactive narrative tooling for building branching stories and experimenting with structured fiction.

Because writing a story and designing the system containing the story are apparently two different hobbies I decided I needed.

A growing collection of creative Google Docs templates built by abusing tables, borders, spacing, and every other layout trick Google Docs reluctantly permits.

Because Google Docs is powerful.

It is also beige in spirit.

Obsidian themes for people who have never once asked:

“Could we make this more subtle?”

Hot pink. Electric lime. Cyan. Ultraviolet. Hard edges. Loud hierarchy.

Nothing I make is quiet.

• I like archives.

• I like taxonomies.

• I like metadata.

• I like taking enormous piles of information and figuring out how they relate to each other.

A surprising amount of my software is ultimately some variation of:

“There is useful information in this mess. Give me a better way to see it.”

That applies to email archives, personal datasets, repositories, documents, stories, resources, and occasionally my own life.

I'm deeply suspicious of software that only exists for you as long as a company continues giving you permission to use it.

• Platforms disappear.

• APIs close.

• Accounts get banned.

• Features vanish.

• Subscriptions increase.

• Export buttons produce a ZIP file containing seventeen JSON files and the vague suggestion that you go fuck yourself.

So I gravitate toward things that are:

local-first · portable · inspectable · remixable · exportable · understandable

• I like software you can download.

• I like plain files.

• I like tools that don't require accounts.

• I like projects that continue existing after their creator gets bored.

If the data is yours, you should be able to actually use it.

Frontend HTML · CSS · JavaScript · TypeScript · React · Vite

Data & tools Python · Streamlit · SQLite

Making things Obsidian · GitHub · Cursor · GIMP · FFmpeg

And whatever else the current project forces me to learn.

find interesting problem
        ↓
become disproportionately curious about it
        ↓
collect entirely too much information
        ↓
invent classification system
        ↓
build tool
        ↓
redesign tool
        ↓
add unnecessary but delightful feature
        ↓
publish to GitHub
        ↓
notice another problem
        ↓
repeat