AI-Powered Packaging Lead Generation & Outreach Agent Overview We are looking for a developer to build an AI agent pipeline that automatically identifies, qualifies, and reaches out to potential customers in the hospitality and food industry based on their packaging usage. The agent will research businesses, analyse their packaging via website menus and images, score them as leads, and send personalised cold emails. Business Context We are a packaging supplier that sources products direct from manufacturing partners in China and stores them locally in Australia. We specialise in high-volume consumable lines — primarily packaging and disposables — for businesses in hospitality, food production, bakeries, and retail. Our ideal customer is a business doing consistent, high volume on core packaging lines (e.g. burger boxes, bags, coffee cups, greaseproof paper, snack boxes) who is currently overpaying through a local distributor. What We Need Built A fully automated agent pipeline with the following stages: Stage 1 — Lead Scraping Integrate with a lead scraping tool (e.g. Lobstr.io, Outscraper, or Scrap.io) via API Pull business name, website URL, phone number and email address Target industries: cafes, restaurants, bakeries, fast food, food production, hospitality Target location: Australia (initially Perth, WA) Output: structured list of leads with contact details Stage 2 — Website & Menu Analysis Agent visits each business website Reads and analyses the menu or product pages Identifies what packaging they are likely using based on the food/products they sell Examples: burger boxes, coffee cups, paper bags, greaseproof, snack boxes, takeaway containers Output: list of likely packaging lines per business Stage 3 — Google Images Analysis Agent searches Google Images for each business Uses computer vision (via OpenAI Vision or Claude API) to identify packaging visible in food photos Confirms and adds to the packaging profile built in Stage 2 Output: confirmed packaging types with confidence score Stage 4 — Lead Scoring & Ranking Score each lead based on: Number of core packaging lines identified Estimated volume (based on business size, number of locations, menu size) Match to our target product lines Rank leads from highest to lowest priority Filter out low-volume or poor-fit businesses Output: ranked lead list with score and packaging summary per business Stage 5 — Personalised Email Generation & Sending Generate a personalised cold email for each qualified lead Email should reference the specific packaging lines identified for that business Follow a provided email template (we will supply the copy and structure) Integrate with an email sending platform (e.g. Instantly.ai, Smartlead, or similar) Output: emails sent to ranked leads, results logged Technology Requirements AI Agent Framework: OpenClaw (preferred) — open-source agent framework Vision/LLM: OpenAI Vision API or Claude API for image analysis and content generation Email Platform: Instantly.ai or Smartlead (we have an account — developer to integrate) Scraping Integration: API connection to Lobstr.io or Outscraper Language: Python preferred Hosting: Agent should be able to run on a schedule (daily or weekly) All credentials and API keys will be supplied by us Deliverables Fully functional agent pipeline covering all 5 stages above Simple dashboard or log showing leads scraped, scored, and emailed Ability for us to review and approve leads before emails are sent (optional approval step) Documentation explaining how to run, update, and maintain the agent 2 weeks post-launch support for bug fixes Out of Scope We do not need a CRM built — a simple spreadsheet or Airtable output is fine We do not need a full website or customer portal Social media outreach is not required at this stage Budget & Timeline Please provide a fixed-price quote for the full build Indicative timeline: we would like this live within 4–6 weeks Milestone-based payments preferred (e.g. 50% upfront, 50% on delivery) What We're Looking For in a Developer Experience building AI agents or automation pipelines Familiarity with OpenAI or Claude APIs (vision and text) Experience with web scraping and lead enrichment tools Strong communication — we are not technical so plain English updates are important Portfolio or examples of similar projects highly regarded To Apply Please respond with: Your relevant experience and any similar projects you have built Your fixed price quote and proposed timeline Any questions or clarifications you need before quoting