The same pattern — curate sources, structure them, let AI extract and validate — applied to whole domains of life. Each turns a pile of raw history into something you can actually ask.
The same pattern — curate sources, structure them, let AI extract & validate — applied to whole domains of life. Each turns a pile of raw history into something you can actually ask.
Years of WhatsApp, mail and CRM history → a structured picture of relationships: who, when, what was said, what matters.
Zettelkasten for the AI age — curated sources → AI-extracted knowledge highlights → validated → building blocks for writing & talks.
Sender-centric: classify every sender signal-vs-noise, auto-generate filters, turn an overflowing inbox into a managed source.
A 245k-message newsletter/promo archive mined with DuckDB + AI + rules — what the market is actually saying, over years.
Daily notes + curated photos → a trips database → curated trip data you can share with a travel companion. Live.
The overarching concept binding them: which sources (people, orgs, mail, reading) deserve trust — scored, then routed into the right system.
Different domains, one method. Structure the sources, apply AI with rules, validate the output — and a lifetime of scattered data becomes a set of systems that compound.