MoVoice

"The Conversationalist" — "I'm here to talk training"

Status: ❌ Future

MoVoice will be the AI chat interface for conversational coaching.

---

Purpose

• Natural language interaction
• Answer training questions
• Provide personalized advice
• Explain recommendations
• Adjust plans via conversation

---

Example Conversations

Asking About Progress

User: "How am I doing this week?"

Mo: "Great week so far! Here's your summary:

📊 3 sessions completed
💪 2 new PRs (Squat 275x3, Bench 225x5)
😴 Sleep averaging 7.2 hours
⚡ Energy levels good (4.2 avg)

Your fatigue score is 4 (normal) - you're in a good spot
to push hard in your next session.

Anything specific you want to focus on?"

Asking for Advice

User: "Should I skip my workout today? Feeling tired."

Mo: "Let me check your data...

Your fatigue score is 6 (elevated) and energy is 2 today.
Sleep was only 5 hours last night.

My recommendation: Do a lighter session today.
- Reduce sets by 30%
- Stay at current weights (don't try to progress)
- Focus on movement quality

Or you could take a full rest day - you've trained 4 days
in a row. Sometimes rest IS the workout.

What feels right to you?"

Exercise Questions

User: "What's a good alternative to barbell rows?"

Mo: "For horizontal pulling, here are some great alternatives:

1. **Dumbbell Rows** - Similar movement, more range of motion
2. **Cable Rows** - Constant tension, easier on lower back
3. **T-Bar Rows** - Heavy loading, great for thickness
4. **Chest-Supported Rows** - No lower back involvement

Based on your equipment (full gym) and history,
I'd suggest Cable Rows - you've responded well to
cable work in the past.

Want me to swap it into today's workout?"

---

Data Model

interface ChatMessage {
  id: string;
  sessionId: string;
  role: 'user' | 'assistant';
  content: string;
  timestamp: Date;
  context?: MessageContext;
}

interface MessageContext {
  fatigueScore: number;
  recentWorkouts: WorkoutSummary[];
  relevantPRs: PersonalRecord[];
  activeGoals: Goal[];
}

interface ChatSession {
  id: string;
  userId: string;
  startedAt: Date;
  lastMessageAt: Date;
  messageCount: number;
}

---

Technical Architecture

User Message
      ↓
Context Gathering (user data, recent workouts, goals)
      ↓
Prompt Construction (system prompt + context + message)
      ↓
LLM (Claude/GPT-4)
      ↓
Response Validation
      ↓
Action Extraction (if user wants changes)
      ↓
Response to User

---

Planned API Endpoints

Endpoint	Method	Description
/api/chat	POST	Send message, get response
/api/chat/history	GET	Get conversation history
/api/chat/sessions	GET	List chat sessions

---

Safety Guardrails

• No medical advice
• Redirect injury questions to professionals
• Encourage rest when fatigue is high
• Don't recommend extreme protocols
• Always mention consulting a doctor for health concerns

---

Implementation Tasks

• Design chat data model
• Build context gathering system
• Create prompt engineering framework
• Integrate LLM API
• Build chat UI
• Add action extraction (modify workouts)
• Implement safety guardrails