Hi everyone,
I wanted to share a surveillance resource and get the community’s feedback on whether it’s useful for epi work.
I’ve built Guardian Watch Worldwide, an automated surveillance system that scans 100+ global sources (WHO, CDC, ECDC, Africa CDC, The Lancet, PubMed, and others) and structures the findings into a consistent, analysis-ready dataset. Each signal carries a timestamp, source, pathogen/disease, location, and a priority score, available as CSV. It currently captures ~1,500 signals per week across zoonotic, foodborne, and environmental disease events, all normalized into one schema.
I’m sharing it openly with the epi community in case it’s useful for empirical work — reporting-lag comparisons across agencies, spatiotemporal cluster analysis, NLP-based surveillance text classification, signal-detection evaluation, etc. It can also serve as a base layer to combine with your own data (climate, case counts, genomic, etc.).
I’d genuinely welcome feedback: Is this kind of normalized multi-source signal feed useful for your work? What fields or features would make it more usable for epidemiologists?
It’s an independent, non-commercial research project. Live dashboard: gww-web.onrender.com
Jae Seok Bae, DVM
One Health Analytics