Fourier Research — Evidence-First AI Research with Hybrid Retrieval (RRF)
Investment-grade AI research with zero fabrication. Every chat turn runs hybrid retrieval — vector (pgvector) and keyword (Postgres full-text) search fused via Reciprocal Rank Fusion — so the model only ever sees the highest-signal evidence, and every factual claim is grounded in verified sources with anchored citations.
What Fourier does
- Hybrid retrieval on every turn — vector + keyword search fused via Reciprocal Rank Fusion (RRF, k=60) across your documents, conversations and highlights
- Live web research with verified citations across every claim
- Deep research reports with cross-verification and source grounding
- Document analysis for PDFs, Word documents and spreadsheets
- Multi-format outputs: chat, executive summaries, slides, audio briefings, flashcards
- Project management with task boards, timelines and risk maps
- Google Workspace integration for Calendar, Drive, Docs and Sheets
- Study mode with personalised learning, progress tracking and Socratic feedback
For analysts, researchers and decision-makers
Fourier is built for professionals who need accurate, verifiable information. Every output is traceable to its source. Enable JavaScript to access the application, or learn more at fourieriq.org.