Sachin Bharadwaz | Projects

Problem

NFCC, a non-profit food bank serving North Fulton County, runs a points-based shopping system where clients select goods using a fixed point budget. Pricing those points required staff to manually look up market prices for thousands of products across categories, then translate prices into point values every cycle. The process was slow, inconsistent, and pulled staff time away from direct community work.

Approach

The capstone team designed a Python pipeline to standardize product data and automate the pricing logic. NLP techniques were applied to clean and normalize inconsistent product names so similar items could be grouped reliably. K-Means clustering organized the standardized catalog into coherent product groups, which gave the pricing engine a stable structure to operate on.

The Walmart Pricing API was integrated to pull live market prices in real time. An automation engine then translated those prices into the point values NFCC uses internally, removing the manual lookup step entirely.

Outcome

The system replaced manual pricing work with a repeatable automated process. It saves NFCC approximately $4,500 per pricing cycle and removes roughly 7 hours of manual effort per cycle, freeing staff to focus on operations rather than data entry. The project placed first at the 2025 Georgia Tech Senior Design Expo, competing against capstone teams from across the Industrial Engineering program.

Georgia Tech Research feature →