Automated Supplier Sourcing Tool

A food & beverage company needed to find EU-organic certified suppliers for specific ingredients — think acai, jabuticaba, mango. The issue? Hundreds of PDF certificates scattered across a government directory, each in a different language. Manually reviewing them took weeks per sourcing round.

The system works in two parts:

Pulls supplier metadata from a public certification API

Downloads PDF certificates via headless browser scraping

Parses product listings from multi-language PDFs (handles Portuguese, French, Spanish)

Loads everything into a database with full-text search indexing

One command. ~10 minutes. Hundreds of suppliers, thousands of products — indexed and searchable by the user.

Real-time search with instant results as you type

Filter by country and product category

Results grouped by supplier with matched products highlighted

Direct links to original certification documents for verification

Next.js 14 / TypeScript / Tailwind CSS / Supabase (PostgreSQL) / Vercel / Puppeteer / pdf-parse

Supabase full-text search over Algolia or Elasticsearch — right-sized for the dataset, zero extra cost, built-in PostgreSQL power

Scalable architecture — the web app is just a read layer; all complexity lives in the data pipeline, making it easy to re-run and extend to new product categories

Multi-language PDF parsing — normalized spelling (açaí = acai) so searches work regardless of how the certificate was written

What used to take a sourcing team weeks of manual PDF review is now a 10-second search. The pipeline is re-runnable, so the data stays fresh as certifications update.