NoisyNotes

NoisyNotes is a mobile + web AI app that turns messy recordings into usable transcripts when typical speech-to-text fails: factory floors, construction sites, busy kitchens, ride-alongs, and field research. Users record or upload audio, then select the environment profile (machinery, crowd, wind, vehicle) and speaker count. The app runs noise-robust transcription, diarization (who spoke when), and highlights low-confidence segments with a tap-to-fix workflow. It also generates a short incident/visit summary and extracts action items, but keeps the focus on accuracy and traceability rather than “magic” notes. For teams, it supports shared projects, permissions, and export to DOCX/CSV/SRT. It’s realistic about limits: if audio is truly unusable, it flags it early and suggests capture tips (mic placement, levels) to prevent wasted time.