← projects

Life Explorer

Mapping a life through therapy transcriptions and AI

I had years of therapy notes - transcriptions, personal reflections, life plans. Twenty-one documents spanning career changes, relationships, grief, identity. I fed them all to Claude and asked: what patterns do you see?

This was the first thing I ever built with AI. Not a to-do app, not a chatbot - a mirror. The kind that shows you the threads running through your life that you can't see when you're living them.

What came back was structured data: themes with subcategories, behavioral patterns with strength scores, connections between people and events, mappings to psychology frameworks I'd been reading. Raw text became a navigable map.

From transcriptions to structure

Raw documents

Twenty-one files - therapy session notes, personal reflections, a life planning spreadsheet, free-form journal entries. Unstructured, emotional, spanning years. Some in Portuguese, some in English.

Claude extracts patterns

Each document was analyzed for themes (relationships, career, identity, grief, health, geography, finances, purpose), people mentioned, behavioral patterns, key quotes, and connections to psychology frameworks.

Interactive exploration

The structured data became a single-page app with six different ways to navigate the same life - each revealing something the others don't.

Six ways to explore

1

Discovery

An insights grid - filterable by theme, searchable by keyword. Each insight card expands to reveal the full analysis, related quotes, and callouts connecting to other themes.

2

Constellation

A D3 force-directed graph where themes, people, and events become nodes. Connections emerge as you explore - who appears together, which themes cluster, what pulls toward what.

3

Timeline

Events laid out chronologically. Click any moment to see what themes were active, who was present, and how it connects to what came before and after.

4

Dashboard

A three-panel deep dive - theme tree on the left, event detail in the center, related insights on the right. Pick any theme or subcategory and see everything connected to it.

5

Frameworks

Maps life patterns to thinkers and authors - Peter Attia on health, Sam Harris on purpose, Derek Sivers on career. Shows where different frameworks converge on the same themes.

6

Journal

Write in real-time and watch the system detect themes, surface related events, suggest frameworks, and find insights - all based on what you're typing, connected to the existing data.

What emerges

The surprise wasn't any single insight - it was seeing the connections. How a career decision echoed a relationship pattern. How a recurring fear mapped perfectly onto a framework I'd been reading about but never applied to myself.

The constellation view was the revelation. Themes that felt separate in life - grief and career ambition, body image and independence - turned out to be deeply connected. The graph made invisible threads visible.

Under the hood

HTML CSS JavaScript D3.js Claude localStorage

A single self-contained HTML file - ~8,000 lines of code with embedded data. D3.js powers the constellation graph and timeline. Journal reflections persist in localStorage. No server, no build step, no dependencies beyond D3.

Try it

The version below uses fictional data - a realistic but invented therapy journey. The structure, interactions, and all six exploration modes work exactly as the original.

Open Life Explorer →

What's next

  • Theme extraction from unstructured documents
  • Six exploration modes (Discovery, Constellation, Timeline, Dashboard, Frameworks, Journal)
  • D3 force-directed constellation graph
  • Real-time journal with theme detection
  • Framework mapping to authors and thinkers
  • Import your own documents
  • AI-powered journal analysis
  • Export insights and patterns
  • Cross-session pattern tracking

Active · Started March 2026