ISSR - AI-Powered Funding Intelligence (FOA Ingestion + Semantic Tagging)

This project proposes an automated pipeline for ingesting, structuring, and semantically tagging federal Funding Opportunity Announcements (FOAs). The system aggregates opportunities from multiple sources, including the Grants.gov public API and NSF funding pages, combining API retrieval with web scraping to handle both structured and semi-structured announcements. Legacy FOA documents are processed using layout-aware PDF parsing to preserve reading order and extract key program information. The extracted records are normalised into a standardised JSON schema and enriched using a hybrid semantic tagging framework. Rule-based matching identifies ontology-aligned terms from controlled vocabularies, while sentence embeddings capture semantic relationships for more flexible tagging. The resulting structured dataset enables efficient discovery, search, and future grant-matching workflows.