Brainforge Knowledge

TICKET_EVENT_FOLLOW_UP_AGENT

6 min read

Ticket: Event Follow-Up Agent - Initial Testing & Usage

Purpose: Handoff documentation for team member to test and use the Event Follow-Up Agent Agent: event-follow-up-agent Status: Ready for Testing Created: 2025-01-16

Overview

What This Agent Does: The Event Follow-Up Agent generates hyper-personalized event follow-up email sequences based on contact segmentation, persona, and engagement context. It drafts emails for human-in-the-loop review and send—turning warm event connections into qualified pipeline.

Model: Based on 11x’s Alice event follow-up system, adapted for Brainforge’s human-in-the-loop workflow.

Key Features:

  • Parses event data (attendee lists, booth scans, meeting logs) in any format
  • Segments contacts by engagement type (Booth Visitor, Panelist, VIP Dinner, Meeting Logs, General Attendee)
  • Identifies persona (Executive, Practitioner, VC/Investor) and adjusts tone/messaging
  • Generates hyper-personalized emails with specific event context and engagement details
  • Qualifies leads through natural conversation questions
  • Creates detailed handoff summaries for AEs

Setup Instructions

Step 1: Pull Latest Code

cd /path/to/brainforge-vault
git pull origin main

Step 2: Verify Agent Files Exist

Check that these files exist:

  • gtm/agents/.claude/agents/event-follow-up-agent.md (Agent definition)
  • gtm/agents/memory/event-follow-up-playbook.md (Playbook/knowledge base)
  • gtm/agents/examples/event-follow-up-workflow.md (Example workflow)

Step 3: Set Up Claude Desktop (If Needed)

The agent is configured for Claude Desktop with the @event-follow-up-agent tag.

To use in Claude Desktop:

  1. Make sure Claude Desktop can access the brainforge-vault directory
  2. Invoke the agent with: @event-follow-up-agent

Alternative: Use Cursor/AI Chat

  • If using Cursor or similar IDE, you can invoke the agent directly in chat
  • Reference the agent files in the conversation

How to Use the Agent

Step 1: Prepare Your Event Data

You need:

  • Event name, date, location
  • Contact list with engagement context

Input Format (Any of these work):

  • Raw text/notes
  • List format
  • Meeting logs
  • Structured data (if you have it)

Required Information per Contact:

  • Name, Role, Company
  • Contact Type (Booth Visitor, Panelist, VIP Dinner, Meeting Logs, General Attendee)
  • Engagement Context (what they asked about, discussed, notes from interaction)
  • Email (if available)

Example Input Format:

Event: NRF 2025
Date: January 12-14, 2025
Location: NYC

Contacts:
1. Eric Joo, Head of NA Feedstock Operations at [Company]
   - Contact Type: Booth Visitor
   - Engagement: Asked about AI ops for supply chain logistics
   - Context: Engaged at Booth #118, very interested in automated workflows
   - Notes: Mentioned they're struggling with fragmented data across systems

2. Sofia Russo, GTM/Strategic Sales Advisor at Cresta
   - Contact Type: Panelist
   - Engagement: Presented on AI sales tools, brief chat after panel
   - Context: Discussed data infrastructure needs for AI sales enablement
   - Notes: Executive persona, interested in strategic partnership

Step 2: Invoke the Agent

In Claude Desktop:

@event-follow-up-agent

I need to follow up on [Event Name] contacts. Here's the event data:

[Paste your event data here]

In Cursor/IDE Chat:

I need to generate event follow-up emails. I'm using the event-follow-up-agent.
Here's the event data:

[Paste your event data here]

Step 3: Agent Processes & Generates Drafts

The agent will:

  1. Parse your event data
  2. Segment contacts (Booth Visitor, Panelist, etc.)
  3. Identify persona (Executive, Practitioner, VC)
  4. Check ICP fit
  5. Generate hyper-personalized Touch 1 emails for each contact

Output Format: Each email draft will include:

  • Contact information and segmentation
  • Complete email (Subject + Body)
  • Personalization details (event context, engagement details, company research)
  • Quality checklist
  • Next steps for human review

Step 4: Review & Adjust Drafts

For each email draft:

  1. ✅ Review personalization (specific event context? engagement details? company research?)
  2. ✅ Check tone matches persona (Executive = concise, Practitioner = tactical, VC = vision-led)
  3. ✅ Verify no generic phrases (“Great to meet you!”, “Hope you’re doing well”)
  4. ✅ Ensure clear CTA
  5. ✅ Adjust if needed

What to Look For:

  • Good: Specific event context, specific interaction details, company-specific research, relevant case studies
  • Bad: Generic phrases, vague references, no personalization, sales-y language

Step 5: Send Emails

For each approved email:

  1. Copy the email (Subject + Body)
  2. Paste into your email client (Gmail, Outlook, etc.)
  3. Send
  4. Log in your CRM/tracking sheet

Recommended Tracking:

  • Contact name, company, email
  • Contact type (Booth Visitor, etc.)
  • Persona (Executive, Practitioner, VC)
  • Email sent date
  • Response status

Step 6: Track Responses & Generate Follow-Ups

When contacts reply:

  1. Share the response with the agent:

    @event-follow-up-agent

Contact: [Name] from [Company] replied:
"[Their response text]"

Generate a follow-up response.

  2. Agent will:

    • Analyze the response for qualification signals
    • Generate appropriate follow-up
    • Ask qualification questions naturally (if engaged)
    • Score the lead (if qualified enough)

For highly qualified leads (score 40-50):

  • Agent will generate a handoff summary
  • Route to AE immediately
  • Include conversation flow, qualification signals, next steps

Example Workflow

See gtm/agents/examples/event-follow-up-workflow.md for a complete example with:

  • Sample event data input
  • Agent processing steps
  • Generated email drafts
  • Response tracking
  • Qualification workflow
  • Handoff summary

Testing Checklist

First Test Run:

  • Pull latest code
  • Verify agent files exist
  • Prepare test event data (can use example from workflow.md)
  • Invoke agent with test data
  • Review generated drafts
  • Check personalization quality
  • Verify tone matches persona
  • Test with real event data (when ready)

Quality Checks:

  • Each email includes specific event context
  • Each email includes specific engagement details
  • Each email includes company-specific research (if possible)
  • Tone matches persona (Executive vs. Practitioner vs. VC)
  • No generic phrases
  • Clear CTA included

Troubleshooting

Agent Not Found / Not Responding

Issue: Agent doesn’t respond or isn’t found

Solution:

  1. Verify agent file exists: gtm/agents/.claude/agents/event-follow-up-agent.md
  2. Check Claude Desktop has access to brainforge-vault directory
  3. Try invoking with full path or alternative method
  4. If using Cursor, make sure files are in workspace

Emails Too Generic / Not Personalized

Issue: Generated emails lack personalization

Solution:

  1. Check that event data includes engagement context (what they asked about, discussed)
  2. Verify contact type is specified (Booth Visitor, Panelist, etc.)
  3. Ensure persona is identified (Executive, Practitioner, VC)
  4. Provide more detailed notes in input if available

Agent Missing Context / Knowledge

Issue: Agent doesn’t know Brainforge’s ICP, positioning, etc.

Solution:

  1. Verify knowledge files exist:
    • gtm/agents/memory/qualification-criteria.md (ICP)
    • gtm/agents/memory/ideal-customer.md (ICP profiles)
    • gtm/agents/memory/positioning.md (Messaging)
    • gtm/agents/memory/message-templates.md (Templates)
  2. Check if these files are populated (some may be templates)
  3. If templates are empty, agent will still work but may use generic language

Qualification Scoring Seems Off

Issue: Qualification scores don’t match expectations

Solution:

  1. Review scoring criteria in event-follow-up-playbook.md (Qualification Workflow section)
  2. Scores are 1-10 per dimension (ICP Fit, Engagement, Decision Maker, Pain Point, Budget/Timeline)
  3. Total max is 50 (5 dimensions × 10 each)
  4. Adjust expectations or refine scoring criteria if needed

Next Steps After Testing

Once Testing is Complete:

  1. Collect Feedback:

    • What worked well?
    • What needs improvement?
    • Any missing features?

  2. Refine Based on Results:

    • Adjust templates if tone/messaging is off
    • Add more case studies/resources if needed
    • Update ICP/qualification criteria if gaps found

  3. Continue Development:

    • Add more playbooks (different event types)
    • Integrate with CRM (when ready)
    • Add analytics tracking
    • Build long-tail follow-up automation

Key Files Reference

Agent Definition:

  • gtm/agents/.claude/agents/event-follow-up-agent.md

Knowledge Base:

  • gtm/agents/memory/event-follow-up-playbook.md (Segmentation, personas, templates, sequences)
  • gtm/agents/memory/qualification-criteria.md (ICP, BANT, scoring)
  • gtm/agents/memory/ideal-customer.md (ICP profiles)
  • gtm/agents/memory/positioning.md (Brainforge messaging)
  • gtm/agents/memory/message-templates.md (Email templates)

Examples:

  • gtm/agents/examples/event-follow-up-workflow.md (Complete example workflow)

Questions?

If you encounter issues:

  1. Check this ticket first
  2. Review example workflow for reference
  3. Check agent definition file for capabilities
  4. Ask in team channel or tag [Your Contact]

For Feature Requests or Improvements:

  • Add to ticket or create new ticket
  • Include examples of what didn’t work or what you need

Last Updated: 2025-01-16 Created By: Robert Tseng Testing Status: Ready for Initial Testing

Graph View