-
FNB DataQuest Recommender System
Collaborative Filtering | Python, Pandas, NumPy, Scikit-learn
This project involved building a customized collaborative filtering recommender system for the FNB DataQuest hackathon. Architected similarly to the famous Netflix Prize matrix factorization competition, the engine analyzed historic customer transaction behaviors. Its purpose was to predict and rank the top 5 financial products a user would interact with next.
View on GitHub
-
Harvest
[Active Development] | Web Architecture | Springboot, Java
An application built to bridge the gap in the agricultural supply chain by allowing consumers to purchase produce directly from small and micro-scale farmers. The platform focuses on connecting customers to local growers, enabling small-scale producers to generate income.
View on GitHub
-
Photo Quality Checker
[Active Development] | Data Science | Computer Vision
A data science utility developed to programmatically analyze and determine whether a given photograph meets quality thresholds. The system leverages computer vision and statistical analysis to evaluate photographic attributes, filtering out bad captures based on underlying structural data.
View on GitHub