
Home » Research Database » Data Archives » Academic and Scientific Data » Richard Geldreich Archive
Richard Geldreich, Jr. approaches UAP research through the lens of a systems programmer and data-preservation engineer rather than a traditional historian. A veteran C/C++ graphics and compression specialist, Geldreich is the co-founder of Binomial LLC (creators of the Basis Universal GPU texture compression system) and previously worked as an optimization specialist at SpaceX (Starlink).
That engineering pedigree defines his archival philosophy in the UAP field. Historical UAP data has long been fragmented across out-of-print books, unindexed PDFs, decaying personal websites, and obsolete binary formats from the DOS era. Through his E.C.H.O. (Event Chronology and Historical Overview) Project and his GitHub repositories, Geldreich treats UAP history as a data-normalization and software-longevity problem: rescuing legacy catalogs, converting them into open, machine-readable standards (UTF-8 JSON and Markdown), and deploying zero-maintenance, client-side tools designed to remain functional for decades without backend servers.
ufo-search.com (The E.C.H.O. Project Front End)What It Contains: This website is the public-facing interface for Geldreich’s E.C.H.O. Project. It hosts three main components: the UFO/UAP Event Chronology Search Engine, a unified UFO/UAP Event Timeline, and a Crash Retrieval Conference Proceedings directory with a Key Word In Context (KWIC) index. When opened, it downloads the entire compressed historical database into the user’s browser memory and builds a local inverted search index. It supports word stemming, diacritic normalization (e.g., matching both “Vallee” and “Vallée”), source filtering, and flexible date-range or partial-date queries (such as searching a specific month and day across all years).
Importance to UAP Research: ufo-search.com solves two major problems in UAP investigation: cross-catalog discovery and operational longevity. Researchers can query dozens of disparate historical timelines simultaneously in milliseconds to spot recurring dates, locations, or behavioral patterns—such as Dr. Donald Johnson’s “Same Time Next Year Syndrome”. Because it runs 100% client-side with no server-side search logs, no cookies, and no database backend, it protects researcher privacy while remaining virtually immune to server rot.
[github.com/richgel999/ufo_data](https://github.com/richgel999/ufo_data) (“The Dataset of the Damned”)What It Contains: Subtitle-framed as “The Dataset of the Damned” (a nod to Charles Fort), this repository contains both the raw/converted historical UAP datasets and the open-source (Apache 2.0) C/C++ command-line tool (ufojson) that ingests, normalizes, and merges them into a single master chronology (timeline.md and majestic.json). It unifies over a dozen major historical UAP catalogs into UTF-8 JSON and GitHub Flavored Markdown, including:
Larry Hatch’s *U* Database (U.RND): ~18,000 sighting records rescued from Hatch’s 1994–2002 DOS binary program (with assistance from David Marler of the National UFO Historical Records Center).
Major Chronologies & Books: George M. Eberhart’s UFOs and Intelligence, Jacques Vallée’s Passport to Magonia, Richard Dolan’s UFOs and the National Security State appendix, NICAP records, Dr. Donald A. Johnson’s catalog, LeRoy Pea’s chronology, Thomas R. Adams’s The Choppers, Fred Keziah’s 1958 poster, and George D. Fawcett’s 1961 manuscript.
European & Public Domain Sources: Jérôme Beau’s French rr0.org chronology, Godelieve Van Overmeire’s French timeline, nuclear test reference dates, and a ~600-event public-domain anonymous PDF chronology from 2023.
Importance to UAP Research: This is arguably Geldreich’s most significant technical contribution to the field. By reverse-engineering and converting legacy formats like Hatch’s DOS binary file into clean UTF-8 JSON, he unlocked decades of painstakingly compiled UAP history for modern software developers, data scientists, and AI/NLP researchers. He also carefully documents the copyright and permission status of each constituent dataset, setting a high standard for ethical UAP data curation.
[github.com/richgel999/uap_resources](https://github.com/richgel999/uap_resources) (Key OSINT UAP Related Materials)What It Contains: Unlike the structured sighting timelines in ufo_data, this repository is a curated Open-Source Intelligence (OSINT) document archive focused on primary-source records, OCR-processed scans, and credential verification for cleared military, intelligence, and scientific personnel. Its README and file directory compile and mirror key evidence surrounding UAP history and alleged crash-retrieval programs, including:
Personnel & Credential Records: Philip J. Corso’s U.S. Army Officer Qualification Record (DA Form 66) and an OCR-searchable version of his manuscript Dawn of a New Age; June Crain’s Wright-Patterson AFB personnel and security clearance files; Dr. Eric W. Davis’s exported LinkedIn resume; Fred L. Crisman’s AEC security clearance files; and verified background dossiers on Jesse Marcel, Luis Elizondo, Hal Puthoff, John Ramirez, James Semivan, and Donald H. Phillips.
Primary Historical & Declassified Documents: A cleaned, searchable PDF of the Admiral Wilson/Davis notes; the 1952 Maj. Gen. Samford Washington, D.C. press conference transcript; the declassified Australian Scientific Intelligence UAP report; Don Berliner’s Blue Book Unknowns (OCR’d); NIDS autopsy film emails; and scans from the Archives for the Unexplained (AFU) (such as Gerald Light and Riley Crabb materials).
Reference Tools & Research: A DOSBox demo image of Larry Hatch’s *U* database, Peter Sturrock’s time-series analysis of UFO catalogs, Geldreich’s Contact in the Desert presentation slides, and archived screenshots of government webpages that were later removed (“memory-holed”).
Importance to UAP Research: In a topic plagued by broken links, hearsay, and unsearchable low-resolution PDFs, uap_resources serves as an anti-link-rot vault. By applying Optical Character Recognition (OCR) to legacy manuscripts and mirroring primary-source personnel records in a version-controlled Git repository, Geldreich makes it possible to quickly verify the credentials and historical claims of key disclosure figures without relying on secondary summaries.
[github.com/richgel999/nuforc_sightings_data](https://github.com/richgel999/nuforc_sightings_data)What It Contains: A Python-based data collection, cleaning, and geocoding pipeline built for the National UFO Reporting Center (NUFORC) database. The project extracts raw witness reports into line-delimited JSON and produces a standardized, enhanced CSV (nuforc_reports.csv). Standardization includes converting dates to ISO 8601 format, normalizing object shapes (e.g., merging “circular” into “circle”), cleaning state/city abbreviations, and appending latitude/longitude coordinates for U.S. and Canadian sightings. Notably, the pipeline merges new pulls with historical snapshots so that reports later removed from NUFORC persist in the archive.
Importance to UAP Research: NUFORC is the largest public self-reporting UAP databank (150,000+ entries), but its raw entries are notoriously inconsistent and its Terms of Service prohibit live scraping. Mirroring this pipeline preserves both the historical methodology for cleaning NUFORC records and the longitudinal integrity of older reports that might otherwise disappear from the live site.
[github.com/richgel999/ufo-reports](https://github.com/richgel999/ufo-reports)What It Contains: A static, pre-cleaned collection of over 80,000 NUFORC sighting reports spanning nearly a century, formatted specifically for quantitative data analysis. It provides two core CSV datasets:
ufo-complete-geocoded-time-normalized.csv (88,874 total records, including unresolved locations and unparsed durations).
ufo-scrubbed-geocode-time-normalized.csv (81,185 fully scrubbed records with zero missing coordinates and valid, positive durations).
The project solves two major headaches in raw NUFORC exports: it uses regular expressions and lookup tables to normalize chaotic human-entered sighting durations into integer seconds, and it separates 20th- and 21st-century records to resolve NUFORC’s ambiguous two-digit year formatting.
Importance to UAP Research: This is the foundational benchmark dataset behind the widely used Kaggle UFO sightings dataset and numerous academic/statistical UAP studies. By keeping a mirror of it alongside his own repositories, Geldreich ensures that researchers working in his ecosystem have immediate, reliable access to a geocoded, time-normalized baseline dataset for mapping and statistical modeling without having to repeat the tedious scrubbing process.
| Resource | Primary Content | Format / Technology | Core Archival Value |
ufo-search.com |
Unified historical UAP search engine, timeline, and Crash Retrieval Conference KWIC index [3.4.1, 3.4.2] | Client-side C++ / WebAssembly (WASM), HTML/JS [1.3.1, 1.3.2] | Zero-maintenance, private, cross-catalog full-text and date-pattern searching [1.3.1, 1.3.2] |
ufo_data |
“Dataset of the Damned” (~18k Hatch *U* records + 12+ classic UAP chronologies) [1.1.1, 1.4.1] |
C/C++ converter (ufojson), UTF-8 JSON, Markdown [1.1.1, 1.4.1] |
Rescues legacy DOS/print catalogs into a unified, machine-readable standard [1.1.1, 1.4.1] |
uap_resources |
OSINT vault of declassified docs, OCR’d books/memos, and cleared personnel records [1.2.1] | OCR’d PDFs, PNGs, Markdown, Git archive [1.2.1] | Prevents link rot/”memory-holing” and grounds disclosure claims in primary documents [1.2.1] |
nuforc_sightings_data |
NUFORC extraction, standardization, and geocoding pipeline (fork of timothyrenner) [3.2.1] |
Python (Scrapy/DVC), JSON, ISO 8601 CSV [3.2.1] | Preserves longitudinal NUFORC records and standardizes messy sighting fields [3.2.1] |
ufo-reports |
80,000+ geocoded, duration-normalized NUFORC sightings (fork of planetsig) [1.4.2, 3.3.1] |
Complete & scrubbed CSVs, regex/lookup tables [3.3.1] | Ready-to-use benchmark dataset for statistical analysis and geospatial mapping [1.4.2, 3.3.1] |
Additional Geldreich Links:
Through A Glass Darkly Podcast – one minute clip in which Geldreich summarizes his personal hypothesis on the phenomenon
Through A Glass Darkly Podcast – Episode 370: You’ll Never Believe What AI Has Discovered About Roswell | Richard Geldreich