Real-time Demand Sensing MVP

Project Overview

This project was developed for the E-SAN PMU-B Coding & AI Competition, where my team (“Team Orange Cat”) won the Data Science Track.

The Solution

We built a Minimum Viable Product (MVP) for real-time demand sensing to help businesses anticipate market fluctuations.

Methodology

• Data Integration: We utilized Python web crawlers and API integrations to fetch diverse open data, including:
  • Weather patterns
  • Demographic shifts
  • Economic indicators
• Modeling: We trained an XGBoost model to analyze these external factors alongside internal sales data.

Results

The tool provided actionable insights for strategic planning, allowing businesses to adjust their supply chain and inventory in response to real-time signals.

Technologies Used

• XGBoost
• Python (Web Scraping, APIs)
• Data Engineering Pipelines