OpenClaw AI Lead-Generation Agent

Project Brief — OpenClaw AI Lead-Generation Agent

Project Overview

I want an AI assistant built using OpenClaw (https://openclaw.ai) that will automatically:
1. Discover and qualify potential packaging customers in Perth, Western Australia
2. Analyse online assets (Google reviews/photos, menus, social media)
3. Extract business contacts and decision-maker names
4. Generate personalised outreach emails and send them

This agent should run autonomously once configured and be manageable for future updates.

⸻

Core Responsibilities

1. Lead Discovery
• Use OpenClaw to call relevant data sources or integrate with APIs/automation for finding hospitality and food venues in Perth.
• Target categories:
• Restaurants, cafés, takeaway food
• Fish & chip shops, pizza, burger joints, catering
• Data sources may include Google Places API, web search, social listings, etc.

Outcome:
A feed of business profiles with URLs/location.

⸻

2. Google Reviews & Photo Analysis
• Visit Google Reviews or related listing pages
• Scan review photos
• Use image recognition (OpenAI/Claude-like vision models) to detect packaging usage
• Score businesses by packaging relevance (plain packaging = high potential)

Outcome:
A rating or tag per lead indicating packaging need.

⸻

3. Menu Reading / Content Analysis
• Crawl business website menus
• Extract food offerings
• Use text analysis to estimate packaging volume or type needs

Outcome:
Structured insights about what the business sells.

⸻

4. Contact Enrichment
• Find an email address and owner/manager first name from:
• Official website
• Facebook
• Instagram
• LinkedIn (optional)
• Prefer first name from official pages over generic “info@” where possible

Outcome:
A CSV or database with:
• Business name
• Email address
• First name
• Website
• Notes/Relevance score

⸻

5. Personalised Email Outreach
• Compose email templates dynamically using:
• First name
• Business insights (menu, packaging review)
• Send emails via a mail API (e.g., Gmail API, SendGrid)
• Include opt-out tracking and basic analytics

Outcome:
Automated outbound email sequences.

⸻

Technical Deliverables

The developer should deliver:

1. OpenClaw Setup
• Installed and running on a server or cloud VM
• Agent configured with:
• Necessary scripts/skills
• Integration with chosen LLM and APIs
• Task scheduling

2. Data Pipeline
• Leads stored in Airtable/Google Sheets/DB
• Export options (CSV etc.)

3. Email System
• Fully configured sending via API
• Logs and status reporting

4. Documentation
• How to run, update, and manage the system
• How to add/remove sources or change outreach templates

5. Optional Dashboard
• Simple dashboard or UI (Sheets/Airtable view acceptable)

⸻

APIs/Tools Preferences

(Open to others based on developer recommendation)
• OpenClaw (framework)
• Vision model (OpenAI/Claude/Whisper for photos)
• Web scraping (compliant and within terms)
• Contact extraction API (Clearbit, Hunter, Snov.io optional)
• Email API (Gmail, SendGrid, Mailgun)

⸻

Data Compliance Requirements
• Follow relevant terms of use for Google APIs and websites
• Respect anti-spam and email compliance (CAN-SPAM / ACMA-equivalent)
• Throttling and handling rate limits

⸻