A web service linking the owner of a lost pet with whoever spots it. One agent reads images, another predicts behaviour, and together they estimate from weather, time of day and terrain where the animal is likely to be, drawing it as a heat map. Finders' photos are auto-matched by similarity, and both sides can chat directly.
Judges praised the scientific grounding of the animal-behaviour analysis and the way it became a working service, as well as the technical strength behind features such as the heat map. As pet owners, they added that they would be glad to have it.
Splitting the work into two agents, one reading images and one predicting behaviour, gives the team a structure they can talk through and an answer to why a given estimate came out as it did. A lost pet is emotionally loaded subject matter, but deriving a probability from weather, time of day and terrain and drawing it on a map keeps the entry from reading as sentiment alone. Carrying it through to automatic photo matching and a direct owner-finder chat likely helped too.
See all winners of 第3回 AI Agent Hackathon with Google Cloud →
Summaries are written by this site. Project pages include demo videos, screenshots and the team's own write-up (videos may autoplay).