Features and improvements shipping to TOMO
Voice recording is fixed and fully working again. AI is restored across all training functions. And now, if voice doesn't process for any reason, TOMO tells you and keeps the form open so you can log your session manually.
Transcription was silently failing since April 21 due to a wrong parameter being sent to the voice recognition service. Your audio was being uploaded correctly, but the transcription step wasn't completing. This is now fixed — voice-to-text works end-to-end again.
Monthly reviews, quarterly assessments, and insight conversations were returning errors due to an AI model configuration issue. All three are now working correctly.
If voice transcription or AI extraction doesn't complete, TOMO now shows a clear message in the Review form: "Voice didn't process — fill in what you remember and save." The form stays fully editable and the Save button always works, so you never lose a session.
Why this matters: The post-training window is short. Voice is TOMO's core feature and it was broken — that's been fixed. But even with reliable voice processing, audio can fail for reasons outside the app (bad signal, background noise). The new fallback means you can always log your session, no matter what.