Ingestion and Semantic Search: ETL to PostgreSQL + Vector DB with Incremental Sync and PDF Embeddings using AI

Build a robust ETL pipeline that ingests EU Funding & Tenders data (the 10k+ results on https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/calls-for-proposals) via the official Search API (https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/support/apis) into PostgreSQL and a vector database (Supabase/pgvector), which supports incremental (delta) sync, processes PDFs (extraction, chunking, embeddings), and exposes Top‑K semantic retrieval with secondary sorting by status and dates using Cursor/AI-Agents.