🕷️️ Job Radar • SCRAPING

upwork.com 🟢 2026-05-24

🔹 B2B Lead Enrichment AI Agent
👤 Client: 🇺🇸 United States Member since 2014-06-16
💰 Price: ****
🚩 Problem: Manual lead research takes 30-40 minutes per lead, requiring automation to reduce this to 60 seconds while filtering out non-ideal customers.
📦 Existing: SmartScout, Apollo.io, TheirStack, Anthropic Claude API, Google Sheets

Specifications:

[Target] Amazon sellers via SmartScout (category, country, revenue, % FBA)
[Method] Multi-signal software detection (TheirStack API, LinkedIn job posts, careers page scraping, Google Search, LinkedIn profiles)
[Method] Finance contact extraction (Apollo API: Controllers, CFOs, VP Finance, Finance Directors, AP/AR Managers)
[Method] State business registry data extraction (address, registered agent, officers, filing status, formation date, DBAs)
[Method] LLM-based scoring and reasoning (Claude API: fit_score, fit_verdict, software_guess, software_confidence, reasoning, opening_hook, case study reference)
[Security] Centralized API key management
[Security] Early-reject filter (Non-US, sub-revenue threshold, NetSuite users)
[Stack] n8n, Make.com, or Python
[Stack] Google Sheets API
[Format] Google Sheet with three tabs: Ready to Call, Needs Review, Rejected
[UI/UX] Editable config file or simple UI for SmartScout filters and Claude prompts

Workflow:

1. Extract Amazon seller data from SmartScout based on configurable filters.
2. Perform business identity confirmation and software detection via multi-signal scraping/APIs.
3. Apply early-reject filter to drop non-US, low-revenue, or NetSuite users.
4. Enrich remaining leads with Apollo finance contacts and state registry data.
5. Process data through Claude API for scoring, verdict, and sales hooks.
6. Populate Google Sheet tabs based on confidence and fit scores.