I built an automatic food-recall tracker that collects recalls from sources including the FDA, USDA, FSA, RASFF, and others. It currently covers the US, UK, Europe, Canada, and South Africa.

It uses ML to classify recalls by severity, cause, and category, and also tries to detect anomalies/outbreaks and identify similar recalls.

You can sign up for daily email alerts based on your preferences, so you can be notified when a recall is relevant to you (The emails are best-effort and shouldn't be treated as an authoritative source of safety information). My goal is simply to make it easier to keep track of recalls that might otherwise be missed.

I'd be particularly interested in feedback on the classification/anomaly detection, and other any other useful recall information could be added.