July 2023 • Google Cloud Hackathon
July 2023 2 min read
Developed during the GCP Hackathon 2023, this system delivers an intelligent time-series forecasting application to solve hybrid workplace challenges. By predicting office attendance, cafeteria food demand, and desk utilization, organizations can optimize operational costs and enhance employee experience.
Employs Facebook Prophet models (prophet_model.json) trained on historical attendance and seasonal trends to accurately project future workplace resource demands.
Utilizes GCP cloud services for scalable data processing, containerized deployment (Dockerfile), and reliable ML pipeline execution.
Provides a clean user dashboard (app.py) enabling facility managers to simulate scenarios, adjust forecast windows, and visualize peak days.
Includes thorough Exploratory Data Analysis (EDA) notebooks (timeseries.ipynb, eda.ipynb) for feature engineering and trend evaluation.
All illustrations on this website are my own work and are subject to copyright.
All illustrations on this website are my own work and are subject to copyright.