Christian Granados

Data Scientist @Stantec

Walnut Creek, CA, US

Developed XGBoost model, that would be placed in a larger ensemble, for WWTP influent flow prediction using PostgreSQL data querying, Prefect pipeline automation, and nested cross-validation with integrated feature selection and hyperparameter tuning. StreamLit dashboard built for stakeholder reporting and experiment tracking.• Built automated BODR (Basis of Design Report) prototype using Power Apps and Plotly Dash to auto-generate building standard codes; this reduced manual effort by 8-16 hours per project ($2,500 cost savings).• Conducted User Acceptance Testing and analyzed feedback data to standardize project protocols. This will enable future NLP-based analytics development for enhanced managerial oversight over all Stantec projects.