External Publication
Visit Post

To All Animal Lovers and Tech Enthusiasts: Help Us Build a Digital Safety Net! 🐾

Hugging Face Forums [Unofficial] March 23, 2026
Source
This is a really meaningful idea, and I think the best way to make it achievable is to start with a simple first version. You probably do not need a full “AI app” right away, and you likely do not need things like a chat model or OCR as the first priority. For an MVP, you would mainly need: * A simple mobile-friendly app or website where someone can upload a photo, location, and notes * A backend system to save and organize reports * A database to keep each animal’s history/timeline * Photo storage * Map/location support * User accounts for volunteers, admins, vets, and pet owners * Notifications/alerts for lost and found cases * An admin/review panel so reports can be checked, corrected, merged, or removed if they are spam/ duplicates For the AI part, the most realistic first step is not “fully automatic identification,” but: * A photo-matching system that suggests similar animals from past reports * Human review before confirming that two sightings are actually the same animal The biggest hidden challenge will probably be data: * collecting enough animal photos * confirming when two photos really are the same animal * testing how often the matching is right or wrong So a realistic first version could be: * someone uploads a photo + location + note * the system saves the sighting * the system suggests possible similar previous sightings * a volunteer/admin confirms the match * the animal’s profile and history get updated I think this is very doable if the project starts small and practical. The idea is strong, but the key is to treat the AI as a helper in the beginning, not as something that makes all decisions on its own.

Discussion in the ATmosphere

Loading comments...