Skip to content

[Admin Assistant] Task 11: Voice Briefings (TTS Integration) #193

@jsirish

Description

@jsirish

Overview

Implement voice/TTS briefings for hands-free consumption of admin assistant updates.

Requirements

TTS Integration

Provider: ElevenLabs (via sag or OpenClaw tts tool)

Voice Options:

  • Nova (warm, slightly British, natural)
  • Or custom voice clone
  • Multiple voices for different content types

Voice Briefing Types

1. Morning Briefing (Audio)

  • Duration: 2-3 minutes
  • Format: MP3
  • Delivery: Telegram voice message or email attachment
  • Content: Weather, calendar, tasks, priorities

2. Meeting Reminders (Audio)

  • Duration: 15-30 seconds
  • Format: Voice message
  • Delivery: Telegram
  • Content: Meeting title, time, link

3. End-of-Day Summary (Audio)

  • Duration: 2-3 minutes
  • Format: MP3
  • Delivery: Telegram or podcast feed
  • Content: Completed items, tomorrow preview

4. Urgent Alerts (Audio)

  • Duration: 10-20 seconds
  • Format: Voice message
  • Delivery: Telegram (immediate)
  • Content: Alert type, action needed

Audio Content Structure

Script Template:

"Good morning! It's [Day], [Date].

Currently [Weather conditions], with a high of [temp] today.

You have [X] meetings today:
- At [time], [meeting title]
- At [time], [meeting title]

[X] tasks are due today, [Y] are overdue.

Top priority: [Task name]

You have [X] new emails, [Y] are urgent.

Deals closing this week: [X] totaling $[amount].

That's your briefing. Have a great day!"

Delivery Options

Telegram:

  • Send as voice message
  • Auto-play on mobile
  • Downloadable

Email:

  • MP3 attachment
  • Embedded audio player
  • Transcript in body

Podcast Feed:

  • Private RSS feed
  • Subscribe in podcast app
  • Auto-download new episodes

Smart Speaker:

  • Alexa skill (future)
  • Google Home action (future)
  • "What's my briefing?" command

Configuration

voice:
  enabled: true
  provider: elevenlabs
  voice: "nova"
  speed: 1.0
  format: "mp3"
  
briefings:
  morning:
    audio: true
    duration_max: "3m"
  meeting_reminders:
    audio: true
  evening:
    audio: true
    
delivery:
  telegram: true
  email: true
  podcast_feed: false

Deliverables

  • TTS integration script
  • Script templates
  • Audio generation
  • Telegram delivery
  • Email attachment
  • Podcast feed (optional)
  • Transcript generation

Acceptance Criteria

  • Voice sounds natural
  • Audio clear and audible
  • Scripts concise (<3 min)
  • Delivery reliable
  • Transcripts accurate
  • Configurable voice/speed

Technical Notes

  • Use OpenClaw tts tool
  • Cache generated audio
  • Provide text transcript as fallback
  • Respect quiet hours (no audio at night)
  • Test on mobile devices

Relates to #182

Metadata

Metadata

Assignees

No one assigned

    Labels

    autopipephase:readyAutoPipe: Plan approved, ready for builder agent

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions