Extract Fiji Infrastructure Vendor Directory via AIIB PPMS Scraping

Intel / Context Summary

The Asian Infrastructure Investment Bank has provided Fiji's Ministry of Public Works with a Project Preparation Special Fund grant for 12-24 months of feasibility studies, engineering designs, and capacity building for climate-resilient rural infrastructure, but the ministry lacks independent capital and procurement authority to execute the actual construction, creating a critical dependency phase where control of preparation determines control of eventual construction.

Catalyst: Why Now

AIIB's PPMS contains public records of all procurement transactions for Fiji projects, including vendor names, contract values, and project types - but this data is buried in individual PDFs and HTML pages rather than a searchable database, creating an information asymmetry between those who can systematically extract it and those who cannot.

Friction: The Bottleneck

• Vulnerability: AIIB's PPMS contains public records of all procurement transactions for Fiji projects, including vendor names, contract values, and project types - but this data is buried in individual PDFs and HTML pages rather than a searchable database, creating an information asymmetry between those who can systematically extract it and those who cannot.
• Capital yield: Sell dataset for $499-$999 to 20-30 engineering firms = $10k-$30k
• Resource capture: Proprietary Fiji infrastructure vendor database with historical contract patterns
• Sovereignty yield: First-mover intelligence on Fiji's AIIB procurement ecosystem
• Required vectors: Vector: Web Scraping & Data Extraction

Wedge: Execution Protocol

Phase 1: PPMS Structure Mapping & Scraper Development: Manually navigate AIIB's PPMS for Fiji-related projects from the last 3 years. Document the URL patterns, pagination structure, and data fields available (vendor name, contract amount, project ID, procurement method). Build a Python scraper using requests/BeautifulSoup that systematically extracts all Fiji procurement records. → Phase 2: Systematic Data Extraction & Cleaning: Run the scraper to extract all Fiji-related procurement records from PPMS (estimated 100-300 records). Clean the data: standardize vendor names, convert currency amounts to USD, categorize by project type (roads, water, sanitation). Export to CSV with columns: vendor_name, contract_amount_usd, project_id, procurement_method, award_date, implementing_entity (FRA/WAF).

Routing Vectors

Specific Roles Required

Claim Protocol