Example: Meeting Preparation Workflow

A complete walkthrough of using Digital Brain for meeting preparation.

Scenario

You have a call with "Sarah Chen" in 30 minutes and need a quick brief.

Step 1: Agent Activates Skill

User Input: "Prepare me for my call with Sarah Chen"

Agent Recognition:

• Trigger: "prepare for" + person name → meeting prep
• Action: Load Digital Brain skill, network module

Step 2: Contact Lookup

Agent searches network/contacts.jsonl for "Sarah Chen":

{
  "id": "contact_sarahchen",
  "name": "Sarah Chen",
  "handle": "@sarahchen_ai",
  "company": "AI Ventures",
  "role": "Partner",
  "circle": "active",
  "how_met": "She reached out after my thread on agent systems",
  "relationship": "potential_investor",
  "topics": ["ai_agents", "investing", "founder_market_fit"],
  "can_help_with": ["Funding intros", "GTM strategy advice"],
  "you_can_help_with": ["Technical due diligence", "Agent architecture insights"],
  "notes": "Very technical for a VC. Former ML engineer at Google. Values depth over polish.",
  "last_contact": "2024-12-15T00:00:00Z",
  "links": {
    "twitter": "https://twitter.com/sarahchen_ai",
    "linkedin": "https://linkedin.com/in/sarahchen"
  }
}

Step 3: Interaction History

Agent filters network/interactions.jsonl by contact_id:

// Most recent interaction:
{
  "id": "int_20241215_100000",
  "date": "2024-12-15T10:00:00Z",
  "contact_id": "contact_sarahchen",
  "type": "call",
  "context": "Initial intro call. She's interested in our agent framework approach.",
  "key_points": [
    "She liked our context engineering approach",
    "Asked about moat - I explained skill composability",
    "She mentioned her portfolio company BuildAI might need our help"
  ],
  "follow_ups": [
    "Send deck with technical architecture",
    "Intro to BuildAI team"
  ],
  "sentiment": "positive"
}

// Earlier interaction:
{
  "id": "int_20241201_140000",
  "date": "2024-12-01T14:00:00Z",
  "contact_id": "contact_sarahchen",
  "type": "dm",
  "context": "She DMed after my agent systems thread",
  "key_points": ["Impressed by technical depth", "Asked about investment stage"],
  "follow_ups": ["Schedule intro call"],
  "sentiment": "positive"
}

Step 4: Check Pending Items

Agent scans operations/todos.md for Sarah-related items:

### Waiting On
| Task | Waiting On | Since | Follow Up |
|------|------------|-------|-----------|
| BuildAI intro | Sarah Chen | 2024-12-15 | 2024-12-29 |

### P1 - This Week
- [ ] Send technical deck to Sarah Chen

Insight: Follow-up items are overdue!

Step 5: Generate Brief

Agent compiles contextual brief:

# Meeting Brief: Sarah Chen

**When**: Today, [TIME]
**Who**: Sarah Chen, Partner @ AI Ventures
**Relationship**: Potential Investor (active circle)
**Days since last contact**: 14

---

## Quick Context
- Former ML engineer at Google (technical background)
- Reached out after your agent systems thread
- Values depth over polish

---

## Last Conversation (Dec 15)
- Initial intro call - positive sentiment
- She liked context engineering approach
- Asked about moat → explained skill composability
- Mentioned BuildAI (portfolio company) might need help

---

## Pending Follow-ups ⚠️
1. **Send technical deck** - Still pending (P1)
2. **BuildAI intro** - Waiting on her, follow up today

---

## Her Interests
- Agent architecture
- Context engineering
- Technical due diligence

---

## Value Exchange
- **She can help**: Funding intros, GTM strategy
- **You can help**: Technical DD, agent architecture

---

## Suggested Talking Points
1. Apologize for delayed deck, offer to walk through live
2. Ask about BuildAI intro status
3. Share recent progress on [relevant feature]
4. Ask about her investment thesis timeline

---

## Links
- Twitter: @sarahchen_ai
- LinkedIn: linkedin.com/in/sarahchen

Step 6: Post-Meeting Logging

After the call, agent helps log to network/interactions.jsonl:

User: "Just finished with Sarah. She's moving forward with intro to BuildAI and wants to see the deck by Friday."

Agent appends:

{
  "id": "int_20241229_143000",
  "date": "2024-12-29T14:30:00Z",
  "contact_id": "contact_sarahchen",
  "type": "call",
  "context": "Follow-up call. Positive momentum on both fronts.",
  "key_points": [
    "She's proceeding with BuildAI intro",
    "Wants technical deck by Friday",
    "Mentioned Series A timeline is Q1"
  ],
  "follow_ups": [
    "Send deck by Friday Dec 31",
    "Expect BuildAI intro next week"
  ],
  "sentiment": "positive"
}

Agent updates operations/todos.md:

### P0 - Must Do Today
- [ ] Send technical deck to Sarah Chen (due: Friday)

### Waiting On
| Task | Waiting On | Since | Follow Up |
|------|------------|-------|-----------|
| BuildAI intro | Sarah Chen | 2024-12-29 | 2025-01-05 |

Agent updates contact's last_contact:

{
  "id": "contact_sarahchen",
  ...
  "last_contact": "2024-12-29T14:30:00Z"
}

Files Accessed

Total: ~390 tokens

Key Takeaways

1. Context Assembly: Agent pulls from multiple sources
2. Actionable Output: Brief highlights what to discuss
3. Follow-up Tracking: Pending items surfaced proactively
4. Complete Loop: Meeting → Log → Update todos → Update contact
5. Relationship Intelligence: Notes inform meeting approach