Sentiment-aware demand forecasting
I built a sales forecasting system that combined social sentiment scraping with predictive modeling and a dashboard for interpreting demand signals. It was built in a 36-hour sprint and shaped around one question: can noisy public language help forecast what people will buy next?
Role
Builder
System Surface
Sentiment, forecasts, dashboard
Stack
React, Chart.js, Python
Delivery Model
Hackathon sprint
The final system shipped as a working forecasting dashboard and placed second at HackAI 2025. The useful part was not just the model score; it was the translation layer between social text, predictive output, and a dashboard someone could actually use.
Event
HackAI 2025
Project Status
Second place
Primary Focus
Forecast interpretation
Output Format
Dashboard and model