Developer
Won 1st place in the urban health track of the Urban Climate Adaptation Hackathon (ARIES / CESAR School / Recife City Hall), built in 3 days with a team of 6. Pulso predicts arbovirus cases per neighborhood before the outbreak with place, date, and count, recorded immutably so every forecast can be verified later. Built using the XGBoost model combining rainfall, standing-water persistence, SINAN records, and population density into a composite territorial risk index. The judges singled out how deployable it is for public administration: cheap, fast to operate, ready to scale.
Python · XGBoost · Django · PostgreSQL · Railway