WAY Judge - Talk & Reflect

WAY Judge helps you reflect on what was said, not decide who won.

FOUR LANGUAGES Choose English, Korean, Simplified Chinese or Japanese for the interface and speech recognition. Mixed-language auto-detection is not supported.

LISTEN IN REAL TIME With everyone’s consent, transcribe a conversation on your iPhone. See a suggested topic, relative microphone levels and expressions worth reviewing. Select speaker A or B yourself; speaker identity is not detected automatically.

BRING AN AUDIO FILE Select an existing m4a, mp3, wav or caf recording, up to 60 minutes, and transcribe it on-device. The original file is not copied, stored or uploaded. Pauses split the transcript into draft turns; you confirm the speakers.

TAKE A SMALL PAUSE Sustained loud input or potentially harsh expressions can pause listening and offer gentle breathing guidance. These are prompts to rest, not emotion diagnoses.

REFLECT WITH EVIDENCE Review each person’s expressed requests, claims and explanations, with quoted examples of possible profanity, personal attacks, generalizations and past-event references. Get suggestions for the next conversation. Rules can miss context, dialect and quoted speech.

KEEP PEOPLE FOR NEXT TIME Register names or nicknames after a conversation and reuse them later.

OPTIONAL AI, YOUR OWN API KEY Use your commercial OpenAI or Anthropic API key for an additional interpretation. Before every request, review the named recipient and exact data, then separately confirm your consent and every participant’s permission. Only the selected transcript, speaker pseudonyms, language/model and review instructions are sent, not loudness metadata, timestamps, other history or raw audio. Text may still identify people. Provider retention and processing terms apply. Gemini transfers are paused in this release; existing keys are preserved. API fees are separate from the app price and consumer AI subscriptions.

LOCAL BY DEFAULT Transcripts and reports stay in app-private storage, excluded from iCloud and device backups. No raw-audio storage, voiceprints, ads, tracking or WAY server. Preview a name-masked report before sharing and delete records at any time.

Speech features require on-device recognition support for the chosen language on your iPhone. There is no server-transcription fallback. Transcription, overlapping voices and microphone distance can affect results. This app does not determine truth, blame, mental health or a winner, and is not a substitute for counseling, legal advice or emergency services. If there is violence or a threat, stop and seek local emergency help.

Sygnały wykrywalności