THE FOMC ORACLE

Federal Reserve Persona Engine — 1970–Present — Seven Chairs — Corpus-Grounded
📄 PDF Corpus Loader ACTIVE
Paste or load the extracted text from your FOMC PDFs to ground each persona in actual primary source documents — transcripts, minutes, and speeches. The preprocessor script converts your raw files to chair-labeled JSON automatically.
Step 1 — Run the corpus scrapers to download PDFs:
python fomc_corpus_builder.py  (historical, 1976–2015)
python fomc_calendars_scraper.py  (recent, 2015–present)

Step 2 — Run the preprocessor to map docs → chairs:
python fomc_corpus_prep.py --corpus-dir fomc_corpus/texts --merge-dirs fomc_calendars/texts
→ Outputs fomc_corpus_by_chair.json

Step 3 — Load the JSON below, or paste text per-chair manually.
Press Conference Query
Reporter's question
Current Chair's actual answer
Select personas → convene the panel
Current Chair (Powell) — Actual Response
Historical Consensus Signal Computing…
Hawkish (tighten)Dovish (ease)

Convene the panel to hear from history's Fed Chairs.

When History Rhymes
Historical analogs — Lessons from the archive — 1936–present
Searching the archive…