You talk. It listens, remembers, organizes, and asks. You stay the author from the first sentence to the finished work.
A full-screen recorder, an honest timer, and one line: I'm listening. Nothing rewrites itself in front of you while you're thinking. Your audio is written to your device first, then uploaded — if your connection drops, recording keeps going.
People, places, events, themes, threads, and facts are pulled from what you actually said. Your original audio and verbatim transcript are never altered by anything that happens later.
Every stored detail knows where it came from: which session, which moment, and how sure you were. "I think I was around nine" is stored as approximately nine. It never quietly becomes exactly nine.
When something important is missing, LENEGIS asks you — it does not fill the gap with invented facts. Questions are ranked, deduplicated, and never repeated once you've answered or dismissed them.
If you said Marcus moved away in 2003 and a later scene puts him at home in 2004, you'll see a possible conflict — with the option to say it's intentional, once, and never be asked again.
Structured chapters and sections, linked back to the sessions they came from. Read it, talk about it, edit it, ask about it, or see its history. No AI labels. No watermarks. No generated-by footers — ever, including in exports.
