🤖 AI Lead Qualification System

Built by Cecilia Wangui Mutero | AI Automation Engineer

📸 Screenshots

n8n Workflow Canvas

Gmail Notification

Slack Alert

HubSpot Contact Auto-Created

📌 What This System Does

A fully automated, end-to-end AI-powered lead qualification pipeline that:

• ✅ Captures inbound leads from a Tally web form via webhook
• ✅ Uses Groq LLaMA 3.1 AI to score and qualify each lead (1–100)
• ✅ Routes qualified leads to HubSpot CRM automatically
• ✅ Sends instant Gmail email notifications with full lead details
• ✅ Posts real-time Slack alerts to the #qualified-leads channel
• ✅ Silently drops unqualified leads with zero manual intervention
• ✅ Runs 24/7 on autopilot — zero human involvement required

All tools used are on free tiers. Total infrastructure cost: $0.

---

🏗️ Architecture Overview

┌─────────────────┐
│   Tally Form    │  ← Lead submits inbound form
└────────┬────────┘
         │ POST webhook
         ▼
┌─────────────────┐
│ n8n Webhook     │  ← Receives and parses form data instantly
│ (Trigger Node)  │
└────────┬────────┘
         │
         ▼
┌─────────────────┐
│  Groq LLaMA 3.1 │  ← AI scores lead 1-100 based on need + budget
│  AI Qualifier   │     Returns: { status, lead_score }
└────────┬────────┘
         │
         ▼
┌─────────────────┐
│  Parse + Extract│  ← Cleans AI response, extracts contact fields
│  (Code Node)    │
└────────┬────────┘
         │
         ▼
┌─────────────────────────────────┐
│     Check If Qualified          │  ← Routes based on AI decision
│     (If Node)                   │
└──────┬──────────────────┬───────┘
       │ TRUE             │ FALSE
       ▼                  ▼
┌──────────────┐   ┌──────────────┐
│ HubSpot CRM  │   │  Drop Lead   │  ← Silently discarded
│ Create Contact│  │  (No-Op)     │
└──────┬───────┘   └──────────────┘
       │
       ▼
┌──────────────┐
│ Gmail Alert  │  ← Full lead details + AI score emailed instantly
└──────┬───────┘
       │
       ▼
┌──────────────┐
│ Slack Alert  │  ← Team notified in #qualified-leads channel
└──────────────┘

---

🛠️ Tech Stack

Tool	Role	Cost
n8n	Workflow orchestration engine	Free (self-hosted)
Tally	Inbound lead capture form	Free tier
Groq (LLaMA 3.1 8B Instant)	AI lead scoring engine	Free tier
HubSpot	CRM — contact creation	Free tier
Gmail	Email notification delivery	Free
Slack	Team alert channel	Free tier
ngrok	Webhook tunneling (dev)	Free tier

---

🧠 How the AI Qualification Works

The Groq LLaMA 3.1 model acts as a strict qualification engine. It receives:

• The lead's stated need (what they want help with)
• Their declared budget

And returns a clean JSON object:

{
  "status": "qualified",
  "lead_score": 87
}

Qualification rules:

• Budget above $500 + clear business need = qualified
• Vague need or budget below $500 = unqualified
• Score range: 1–100 (higher = stronger lead)

The AI prompt is engineered to return raw JSON only — no markdown, no explanations, no filler text.

---

📋 Prerequisites

• n8n installed (self-hosted or cloud)
• Groq API key (free)
• HubSpot account (free CRM)
• Tally account (free forms)
• Slack workspace (free)
• Gmail account with OAuth2 enabled
• ngrok for local webhook tunneling

---

🚀 Setup Instructions

1. Clone This Repository

git clone https://github.com/CeciliaMutero/ai-lead-qualification-system.git
cd ai-lead-qualification-system

2. Import the n8n Workflow

1. Open your n8n instance at http://localhost:5678
2. Click + → New Workflow
3. Click the three-dot menu → Import from JSON
4. Paste the contents of workflow.json → click Import

3. Configure Credentials

Groq API Key

1. Go to console.groq.com → API Keys → Create
2. In n8n: Groq Lead Qualifier node → Authentication → Header Auth
   • Name: Authorization
   • Value: Bearer YOUR_GROQ_KEY

HubSpot Private App

1. Go to HubSpot → Settings → Integrations → Private Apps → Create
2. Add scopes: crm.objects.contacts.write, crm.objects.contacts.read
3. Copy the token → paste into n8n HubSpot credential

Gmail OAuth2

1. Go to Google Cloud Console
2. Enable Gmail API
3. Create OAuth2 credentials → connect in n8n

Slack Bot Token

1. Go to api.slack.com/apps → Create App
2. Add scopes: chat:write, chat:write.public, channels:read
3. Install to workspace → copy Bot Token (xoxb-...) → paste into n8n

4. Set Up Tally Form

Create a Tally form with exactly these fields:

Field Label	Field Type
Name	Short answer
Email	Email
What do you need help with?	Long answer
Budget	Short answer

5. Connect Tally Webhook

1. Start ngrok: ./ngrok http 5678
2. Copy your ngrok URL (e.g. https://abc123.ngrok-free.app)
3. In Tally → Integrate → Webhooks → Add endpoint:

   https://abc123.ngrok-free.app/webhook/inbound-leads
   

6. Activate the Workflow

1. In n8n, toggle the workflow Active (top right)
2. Submit a test form entry
3. Watch the pipeline execute end-to-end ✅

---

📊 Live Demo Results

✅ Qualified Lead Test

Input:

Name: Hellen Moo
Email: hellenmoo@gmail.com
Need: I need help with my website
Budget: $5000

AI Response:

{ "status": "qualified", "lead_score": 90 }

Outcome:

• ✅ HubSpot contact created automatically
• ✅ Gmail notification sent with full details + AI score
• ✅ Slack alert posted to #qualified-leads

---

❌ Unqualified Lead Test

Input:

Name: Waa Ree
Email: waaree@gmail.com
Need: I need help with hubspot
Budget: $55

AI Response:

{ "status": "unqualified", "lead_score": 25 }

Outcome:

• ❌ No HubSpot contact
• ❌ No Gmail notification
• ❌ No Slack alert
• ✅ Lead silently dropped

---

📁 Repository Structure

ai-lead-qualification-system/
│
├── workflow.json          # Complete n8n workflow (import-ready)
├── README.md              # This file
├── screenshots/
│   ├── n8n-canvas.png     # Full workflow view
│   ├── gmail-alert.png    # Email notification sample
│   ├── slack-alert.png    # Slack message sample
│   └── hubspot-contact.png # CRM contact created
└── docs/
    └── setup-guide.md     # Detailed setup instructions

---

🎬 Portfolio Demo

📹 Watch the full live demo video

The demo shows:

1. Live Tally form submission
2. n8n workflow executing in real-time
3. Gmail notification arriving instantly
4. Slack alert posting to the team channel
5. HubSpot contact automatically created
6. Unqualified lead correctly dropped with zero notifications

---

🔧 Customization Options

This system can be extended to support:

• WhatsApp notifications via Twilio or WhatsApp Cloud API
• Calendar booking via Google Calendar integration
• SMS alerts via Africa's Talking API
• Multi-step AI qualification with follow-up questions
• Lead scoring dashboard via Google Sheets or Airtable
• CRM pipeline automation — move leads through HubSpot deal stages automatically
• Industry-specific scoring — tune AI prompt for clinics, real estate, salons, etc.

---

💼 Built For

This system is ideal for:

• Service businesses (clinics, salons, law firms, real estate agents)
• Freelancers managing multiple client inquiries
• Sales teams who waste time on unqualified prospects
• Startups that need automated lead management without expensive CRM tools

---

👩‍💻 About the Author

Cecilia Wangui Mutero — Full-Stack Engineer & AI Automation Specialist based in Nairobi, Kenya.

Specializing in:

• AI workflow automation (n8n, Groq, LangChain)
• Voice AI agents (Vapi.ai, ElevenLabs)
• Full-stack development (Next.js, FastAPI, Django)
• CRM integrations (HubSpot, Airtable, Google Sheets)

📧 cecilmutero66@gmail.com
🔗 LinkedIn
🐙 GitHub

Available for freelance automation projects and full-time remote roles.

---

📄 License

MIT License — feel free to use, modify, and build on this project.

---

⭐ If this project helped you, please give it a star!