Jaya Jyoti

At Avataar Skincare Technologies, I analyze 10K+ record datasets to uncover workflow inefficiencies and improve process efficiency by about 20%.

I've built Python and SQL data pipelines that reduce manual effort by 30–40%, created performance-tracking automations that speed up decision-making by 25%, and applied regression and classification models to improve forecasting accuracy by about 15%.

I also build end-to-end AI products, including a real-time face-recognition attendance system processing 5,000+ images at about 90% accuracy, a fraud detection model trained on 6.3M+ transactions with a 0.96 ROC-AUC score, and an emotion-driven product recommendation system with about 87% accuracy.

I'm interested in turning data, computer vision, and machine learning into practical systems that improve workflows and user experiences.