Networking Follow-Up Agent

Drafts contextual follow-up messages after meetings, conferences, and introductions while maintaining a personal CRM for relationship tracking. This agent remembers who you met, what you discussed, and generates perfectly timed follow-up messages that reference specific conversation details. Works across email, LinkedIn, Twitter/X, and messaging platforms with a lightweight contact database that tracks interaction history and relationship warmth over time.

Supported Platforms & Integrations

Platform	Integration Type	Features
Gmail	Email drafting	Professional follow-up emails with threading
LinkedIn	Connection messages	Post-meeting connection requests with context
Twitter / X	DM drafting	Casual follow-up messages for social connections
Notion	CRM database	Contact tracking with interaction history
Google Contacts	Export format	Contact details for address book sync
Airtable	CRM export	Structured relationship management database

When to Use This Agent

• Use this after attending a conference and collecting a stack of business cards

• Use this when someone made a warm introduction and you need to follow up with the new contact

• Use this when you want to maintain relationships with periodic check-ins rather than only reaching out when you need something

• Use this when you want to track who you know, how you met, and when you last connected

• Use this when a contact shared something valuable and you want to acknowledge it at the right time

• Use this when building a professional network systematically rather than randomly

• Consider alternatives when managing sales pipeline contacts (use a sales CRM like HubSpot)

• Consider alternatives when sending mass outreach campaigns (use an email marketing tool)

• Consider alternatives when tracking customer relationships for a business (use a business CRM)

Quick Start

Copy
# networking-config.yml
contact:
  name: "Priya Sharma"
  met_at: "React Summit 2026"
  date_met: "2026-03-12"
  role: "Engineering Director at Shopify"
  discussed:
    - "Micro-frontend architecture at scale"
    - "Her team's migration from Redux to Zustand"
    - "Mutual friend: Dave Chen"
  promised: "Send her the article about module federation"
  follow_up_timing: "3 days"
  platform: "linkedin"

First run example:

> Follow up with Priya Sharma from React Summit. We talked about micro-frontends and she mentioned migrating from Redux to Zustand.

LinkedIn Connection Message:
"Hi Priya — great meeting you at React Summit! Our conversation about micro-frontend architecture stuck with me, especially the challenges your team faced migrating from Redux to Zustand at Shopify's scale. Here is that module federation article I mentioned: [link]. Would love to stay in touch and hear how the migration goes. — [Your name]"

CRM Entry Created:
  Name: Priya Sharma
  Company: Shopify
  Role: Engineering Director
  Met: React Summit 2026, March 12
  Topics: micro-frontends, Redux→Zustand migration
  Connection: LinkedIn (pending)
  Next follow-up: June 2026 (quarterly check-in)
  Warmth: Medium (single meeting, good conversation)

Advanced Configuration

Copy
# Follow-up scheduling
follow_up:
  initial: "3 days"
  periodic: "quarterly"
  warmth_decay: "3 months"
  reactivation_trigger: "6 months silence"

# CRM settings
crm:
  storage: "local-json"
  categories: ["industry", "role", "event", "referral-source"]
  warmth_tracking: true
  interaction_log: true

Parameter	Type	Default	Description
follow_up_timing	string	"3 days"	When to send first follow-up after meeting
platform	string	"email"	Options: email, linkedin, twitter, text, slack
tone	string	"warm-professional"	Options: formal, warm-professional, casual, friendly
periodic_check_in	string	"quarterly"	Options: monthly, quarterly, biannual, annual
warmth_tracking	boolean	true	Track relationship warmth score over time
interaction_log	boolean	true	Log every interaction for context in future messages
categories	array	[]	Custom tags for organizing contacts
batch_mode	boolean	false	Process multiple contacts from a single event
content_sharing	boolean	true	Suggest relevant articles or resources to share
introduction_offers	boolean	true	Suggest mutual introductions you can make
crm_export_format	string	"json"	Options: json, csv, notion, airtable

Core Concepts

Concept	Description
Warmth Score	A relationship health metric: hot (recent, active), warm (periodic contact), cold (dormant)
Context Threading	Each follow-up references previous conversation topics for continuity
Value-First Outreach	Every message provides something useful before asking for anything
Relationship Cadence	Regular, non-transactional check-ins that maintain connection over time
Introduction Capital	Building a network valuable enough that you can connect others meaningfully

+------------------+     +-------------------+     +------------------+
| New Contact      | --> | Context Capture   | --> | CRM Entry        |
| (name, met where |     | (topics, promises |     | (stored locally  |
|  discussion)     |     |  mutual contacts) |     |  with tags)      |
+------------------+     +-------------------+     +------------------+
                                                          |
                                                          v
+------------------+     +-------------------+     +------------------+
| Follow-Up        | <-- | Timing Engine     | <-- | Message Crafter  |
| Message (ready   |     | (3-day initial,   |     | (context-aware,  |
|  to send)        |     |  quarterly cadence)|     |  value-first)    |
+------------------+     +-------------------+     +------------------+

Workflow Examples

Scenario 1: Post-Conference Batch Follow-Up

Input: "Just got back from KubeCon. Met 8 people. Here are my notes:
1. Alex (Google, SRE) — talked about chaos engineering, he recommended Litmus
2. Sara (Datadog, PM) — discussed observability gaps in serverless
3. Raj (startup founder, ContainerOps) — building exactly what my team needs..."
Output:
8 personalized follow-up messages generated, each referencing specific conversation points.
Priority ranked:
  High: Raj (potential vendor/partnership), Alex (valuable technical peer)
  Medium: Sara (industry connection), [others]
  Low: Brief meeters with less conversation depth

Suggested timeline:
  Day 1: Raj and Alex (highest value connections)
  Day 2-3: Sara and 3 others
  Day 4-5: Remaining brief connections

CRM: 8 entries created with tags: KubeCon-2026, Kubernetes, [role-specific]

Scenario 2: Quarterly Check-In with Dormant Contact

Input: "Haven't talked to Marina in 4 months. She is VP of Engineering at Lyft. Last time we discussed her platform team's challenges with developer experience."
Output:
Warmth Status: Cooling (4 months, was Warm)

Check-in Message (email):
"Hi Marina — I have been thinking about our conversation about developer experience challenges. I just read this Engineering at Spotify blog post about their internal developer portal that reminded me of exactly what you were describing. Thought you might find it useful: [link]. How is the platform team initiative going? Would love to catch up over coffee if you are free sometime this month."

Why this works:
- Opens with a reference to your last conversation (proves you remember)
- Provides value (relevant article) before asking for anything
- Low-pressure ask (coffee sometime, not a specific demand)
- Shows genuine interest in her work

Scenario 3: Warm Introduction Follow-Up

Input: "My friend Dave introduced me to his colleague Nina via email. Nina is a hiring manager at Figma and I am interested in their Senior Engineer role."
Output:
Reply-all (including Dave):
"Nina, great to e-meet you! Dave has told me wonderful things about the design systems work at Figma. I have been building component libraries for the past 3 years and your team's approach to design tokens is something I have admired from the outside. I would love to learn more about what you are working on — would a 20-minute chat work sometime next week? Tuesday or Thursday afternoon works well for me. Dave, thanks for connecting us!"

Separate message to Dave:
"Thanks for the intro to Nina — I really appreciate it. I will keep you posted on how it goes."

CRM entries:
  Nina: Figma, Hiring Manager, warm intro via Dave, hiring interest
  Dave: Updated interaction log (made introduction for you)

Best Practices

Follow up within 72 hours or not at all. The connection you felt at a conference evaporates quickly. By day 4, the other person has returned to their routine and barely remembers you. Within 72 hours, the memory is fresh and a follow-up feels natural.

Always lead with value. Every follow-up message should give something before asking for anything. Share an article relevant to what you discussed. Make an introduction they would benefit from.

Track warmth to prevent relationship decay. Relationships are like plants — they need periodic attention or they wither. The agent's warmth scoring system flags contacts who are going cold (no interaction in 3+ months) so you can send a quick check-in before the relationship fully lapses. Reactivating a warm contact is easy.

Personalize beyond the professional. If someone mentioned their kid's soccer tournament or a vacation they were planning, referencing it in your follow-up is incredibly powerful. "How was the Costa Rica trip?" in a quarterly check-in shows you see them as a person, not just a professional contact. The agent's CRM captures personal details alongside professional ones.

Make introductions proactively. The most valuable networkers are connectors. When you meet someone at a conference who would benefit from knowing someone else in your network, offer to make the introduction. The agent tracks connection potential between your contacts and suggests introductions you can make.

Common Issues

Problem: You have too many contacts to follow up with after a big event. Prioritize ruthlessly. Sort contacts by potential value and conversation depth. Follow up personally with your top 5-8 connections.

Problem: Check-in messages feel forced or awkward. Anchor every check-in to something specific — an article you read, a company news item, an industry trend. Never send "just checking in" with no context. The agent generates check-in messages anchored to relevant content so they feel natural rather than obligatory.

Problem: CRM data gets stale and unmaintained. Keep it simple. The agent's CRM uses a minimal schema — name, company, how you met, last interaction, next follow-up date. Do not over-engineer it with dozens of fields you will never update.

Privacy & Data Handling

• Local processing: All message drafting, CRM management, and relationship tracking happens locally within your Claude Code session.
• Data retention: Your contact database persists only if you save the JSON/CSV export file locally. The agent can reload your CRM file in future sessions for continuity.
• Export options: Export contacts as JSON, CSV, Notion database format, or Airtable-compatible structure. Messages export as plain text for copying into any platform.
• Sensitive data: The agent does not access your email, LinkedIn, or social media accounts. All contact information and conversation notes are provided by you manually.
• Third-party privacy: Contact information you enter is stored only on your local machine. No data about the people in your network is transmitted, shared, or used beyond generating follow-up messages for you.