On-device transcription for outbound voice messages

#### What would you like to do?
I want my device to transcribe a voice message I record before sending, or immediately after recording, so I can visually review what I just said and have a text reference. This is specifically about **outbound** voice messages I send, not only incoming ones.

#### Why would you like to do it?
This would make voice messaging easier to review, easier to reference later, and more comfortable for moments when I want a written anchor for what I just said. It would also bring the experience closer to the kind of visual reference users get in native voice messaging flows like iMessage.

#### How would you like to achieve it?
An iOS-first implementation could generate a transcript locally on-device when available, then show the text alongside or under the voice message with a tap target to expand or read it. The implementation should stay privacy-preserving and avoid server-side transcription if possible.

#### Have you considered any alternatives?
The current audio-only flow works, but it leaves no quick visual reference for the sender. Server-side transcription would add privacy and infrastructure tradeoffs that are less attractive than on-device transcription.

#### Additional context
This request is for the sender’s own voice message transcript, not just transcription of received voice messages. The goal is to let me quickly see what I said, while keeping the audio message itself intact.

No files or tests are named. Start by locating the iOS outbound voice-message recording and send flow, then compare it with any existing incoming-message transcription entry point. Done means a locally generated transcript is shown alongside or below the outgoing voice message, can be expanded or read, preserves the audio, and avoids server-side transcription.

Written by the indexing model from the issue text.