Research & Publications 

Conventional Applicant Tracking Systems (ATS) tend to be inflexible keyword-matchers. This article describes a new two-step process to design a more refined resume evaluation model based on a small language model (<600M parameters) that is finetuned using GRPO on a custom reward function. We indicate that the RL application presents a critical problem of reward hacking due to the initial experiments of aggressive penalties. We have overcome this challenge by refining the reward function repeatedly. Our resulting GRPO-polished model achieves a final accuracy of 91% on unseen test data.

Recent advancements in smart computing and intelligent decision communication systems have enabled new possibilities for sustainable financial technologies. This chapter introduces a hybrid quantum classical model for credit risk prediction that leverages superposition and entanglement to enhance complex financial data processing. Unlike traditional machine learning models that struggle with class imbalance, nonlinear relationships and high dimensional dependencies, the system combines advanced preprocessing, feature selection and quantum kernel computation within a scalable support vector framework. Trained on 33,000 loan records across 12 borrower attributes, it achieves 94.53% accuracy.

This paper presents a hybrid quantum-classical model for detecting metal surface defects using quantum circuits integrated with Convolutional Neural Networks (CNNs). Leveraging the PennyLane framework, we developed an innovative approach that combines quantum computing techniques with traditional deep learning methods to enhance pattern recognition accuracy for non-destructive testing applications.