One record, used a few different ways.
You put things into Dorima by typing, talking, journaling, adding a task or connecting your calendar. Dorima keeps that information together. Every screen is a view of it, and every AI feature reads the same context, so the advisor, the scenario and the morning paragraph are looking at the same record.
That same context can then be used to plan the day, remember projects and commitments, read a journal entry back to you, ask several advisors the same question, think through a scenario, notice patterns across weeks, keep track of people and important dates, and give you a short summary of yesterday. It can also carry on a chat that already knows what you are working on, finish a habit by writing a few lines, and keep a shelf of books and essays chosen from what you write about.
You shouldn't have to organize the same information over and over.
you put in
- journal entries
- tasks and projects
- habits
- calendar
- people
- directions
- attachments
→
one record
- append-only
- encrypted per user
- isolated per account
context, assembled by code
- your recent entries, and whatever is closest to what you asked
- summaries of the weeks behind them
- which people, which dates
- what you said you were aiming at
→
it reads back
- a paragraph each morning
- a narrative each week
- a read on an entry
- advisors, one question at a time
- scenarios, and what would prove each wrong
- the compass
- a reading shelf
The model writes. It does not decide what is relevant, and it does not act on your behalf.