SEAS
SEAS is a snake recognition system built in three parts: a
snake_recognition module that trains and runs the identification model,
a server that exposes it, and a frontend where users submit a photo or
sighting and get back what the snake likely is.
In a country like Nepal, where snakebites are a real medical emergency and most people can’t tell a cobra from a rat snake, a tool that helps identify a snake quickly matters — identification drives the right first aid and the right antivenom.
The interesting engineering lives in the pipeline: taking an uncontrolled photo from the field, getting it through a classifier, and returning a confident answer through an API that the frontend can rely on.