Using ThoughtDump

Reviewing a Dump

After you record, ThoughtDump transcribes what you said and splits it into individual items. When it's confident, it files an item and moves on. When it detects uncertainty about meaning, timing, or a Project, it flags the item for review. AI can misunderstand; check important details even when an item was organized automatically. Saved items remain available while you review them.

Open and review a Dump

Open a capture from Home or History, or follow an Item’s source link to Dump Review. Review the extracted Items, Needs Review flags, confidence indicators, and source transcript. Confirm or discard a proposed Item; edit its type, title, and details as needed. Confidence is a guide, not a guarantee.

Save status and recovery

Check the save status before closing the page. If an upload has not been saved, keep the page open and use Retry upload when connectivity returns. That recording is still in memory and may not survive closing the page. This differs from server-preserved unfinished captures, which you can reopen from Home or History.

For a saved capture, follow its stage-specific retry: transcription, extraction, or Hands-Free follow-up. Completed work is reused; you do not need to record again. If processing stalls, use Retry. Audio-dependent recovery ends once the retained audio is unavailable; saved transcript work can still be used.

Rescue a capture manually

If a capture has no structured Items, or its follow-up context has expired, choose an Item type, title, details, and optionally an existing Project. Manual rescue retains the complete source transcript. Check the saved result before retrying again.

Email Support whenever you need help. A rough time, the step you reached, and any message you saw can help us start investigating; we may ask for more details.

Replay original audio

Open History or Dump Review to play the original recording and use its playback and seek controls. An unavailable-audio state means the original recording cannot be replayed; the separately retained transcript and Items can still be read and edited. Audio access is private and requires the owner to sign in.

Original audio becomes eligible for scheduled cleanup after seven days, including after access ends. This is not a promise of deletion at an exact second or a verification that the live scheduler has run. Item Read aloud generates speech from Item text and has a Stop control; it is separate from original-audio replay and requires access.

← Back to the ThoughtDump overview