CONVERSATION SOFTWARE BY SOCIO / IN DEVELOPMENT

Live Transcription & Conversation Feedback

Follow the conversation as it happens.

I am building remorse to process live speech, track who is speaking, and provide conversational cues through screens and wearable devices.

See the system working ↓

INSIDE THE PROTOTYPE

From speech to feedback.

These excerpts focus on individual parts of remorse running during IYKYD conversations. Each loop keeps the original timing and links to its source recording.

Speaker activity, incoming text, and conversation prompts show how the system responds as a session develops.

Watch the separate speaker traces as the session continues.

PROTOTYPE / FIELD RECORDING

Separate the speakers.

Diarization tracks which speaker is talking over time. Here, separate activity traces sit alongside the transcript, making changes in participation visible during a live conversation.

IYKYD source / 12:00 ↗

Speaker-presence interface with separate activity timelines for five speakers. — Speaker activity shown over a review interval. Open the image to inspect the interface.

Watch new text arrive and the displayed sequence change.

PROTOTYPE / FIELD RECORDING

Process speech as it arrives.

Incoming words update the live transcript and the on-screen Morse sequence. The view also exposes the microphone source and reported delay, making the path from captured speech to interface output visible.

IYKYD source / 1:28 ↗

The remorse mobile app showing session status, insights, a speaker-labeled transcript, and diarization activity. — The mobile session, from incoming speech to speaker activity.

MOBILE APP / LIVE SESSION

The conversation on your phone.

The mobile view brings connection status, microphone and cue controls, session insights, and the transcript into one screen. Speaker labels connect each passage to the diarization activity shown below.

This view makes the running session available on the phone alongside the desktop interface and feedback experiments.

Open full screenshot ↗

A conversational prompt appears within the running session.

PROTOTYPE / FIELD RECORDING

Turn context into a cue.

The conversation compass places a suggested next step beside the current exchange. This excerpt shows a prompt to leave room for a response or explicitly invite another speaker.

IYKYD source / 3:47 ↗

Watch the session change to Phone Alert and register a haptic signal.

HAPTIC FEEDBACK / LIVE EVENTS

Follow a haptic signal.

The interface switches from a speaker turn to Phone Alert, then records phone haptic signals within the session. This excerpt shows the software side of the feedback path: an event arriving and appearing alongside the conversation. Wearable cues extend that work beyond the screen.

IYKYD source / 3:52 ↗

remorse running on a circular watch on a wrist, in the muted setup state. — On-wrist prototype in its muted setup state.

WEARABLE DEVELOPMENT / 2026

Conversation cues on the wrist.

I am extending remorse to circular watch displays as another interface for conversational feedback. The on-wrist photo documents the prototype's setup screen, including its muted state, hub status, and clear and send controls.

The separate Armed screen shows the message area and guidance for pause-to-send and shake-to-clear interactions.

Circular remorse watch screen in the Armed state, with pause sends and shake clears guidance. — Armed interface state.

DEVELOPMENT THROUGH IYKYD

Tested in real conversations.

IYKYD is where I document remorse in use and explore how small cues can support attention and participation. I am developing the transcription software and wearable feedback through Socio.

IYKYD / episode guide & selected passages ↗

The complete session: See the transcript, speaker timeline, session statistics, and open questions together. View the full interface
The wearable experiment: The follow-up introduction explains adding a watch to the hub and mobile setup. Watch the introduction

All project photos & videos