I never kept up with the filing, so I built a second brain you just talk to. The hard part was privacy.
Disclosure up front: I built this (LabA). It's free during the beta, no credit card, until January 2027; after that I expect under $10 a month.
My notes, to-dos, calendar and inbox lived in four places, and I never kept up with the filing. So LabA is capture first, sort never: type or say "remember the gate code is on the back of the garage door" or "remind me Friday at 3 to call the vet", and it files and tags it. Later you ask "what did I decide about the roof?" and it answers from your notes and shows which ones it used. It also makes a morning plan, briefs you before meetings from what your notes say about the people, and runs checks on a schedule ("every Monday, look through my notes for anything I promised someone").
The part I spent the most time on: notes are encrypted on your device with a key only your devices hold, and my servers store a copy it can't read. But an AI can't think about words it can't read. So when you ask something, your device unlocks just what that one request needs and sends it through my server to the AI service, under my account, not yours, and the AI service doesn't train on it. Encrypted at rest, readable only for the request.
My question for people who already run a system: where's your line? Is that good enough for your notes, or is local-only the only thing you'd trust with them?
Short videos of the real app (made-up people and data): https://sensoryfire.com/learn?ref=reddit-secondbrain
Privacy details: https://sensoryfire.com/privacy?ref=reddit-secondbrain
It's a web app (phone or computer, nothing to install); calendar and email are Google-only for now. Sign up: https://sensoryfire.com/?ref=reddit-secondbrain. There's a 🐞 button for anything that breaks. Blunt is good.