Hit record on your phone. Speak into it. Get a transcript, a searchable entry, and an auto-tagged summary — all processed locally or on your own server. No 'your journal data trains our model.' Just a place to think out loud that you actually own.
Tap record on phone or laptop. Whisper runs locally (or you bring your own OpenAI key if you prefer). Transcript lands in your journal within 30 seconds of stopping the recording.
Each entry gets a short auto-summary + 3-5 tags inferred from content. Find old entries by topic, mood, or date. Optional Claude integration (BYOK) for richer summaries.
SQLite FTS over every transcript. Find 'the thing I said about grandma's recipe' in seconds. Filter by date range, tag, or sentiment.
Optional: mood rating per entry (1-5). Theme frequency chart. See what's been weighing on you this month without digging through 30 entries by hand.
Nothing leaves your machine unless you opt in. No usage telemetry, no 'anonymized training data.' Export the whole archive as Markdown or PDF any time, no lock-in.
PWA you bookmark to your home screen. Record, transcribe, search — no app store involved. Works offline, syncs when you're back online if you run a family server.
Hosted plans are early-access. Self-host today; email tim@dangercorn.net to join the waitlist.
voicejournal is one of 220+ verticals scaffolded from the same template. Each gets a working skeleton for free; individual ones get promoted to full products as demand surfaces.