Kappala Saikumar

Part of a team building an Employee Relationship Management (ERM) app for HRBPs to access comprehensive employee data, manage promotions, recommend courses and mentors, and provide company-level insights. Created a web-based chatbot for ERM-related queries, providing instant access to company data and ongoing development of a multi-level query answering global chatbot. Developing a graph-based enterprise level embedding model for candidate and job recommendations, and built a BERT transformer model for role level prediction, enhancing progression recommendations. Built a Streamlit-based web app in Python to analyze job market trends, skills demand, and load company data into ERM, enhancing data-driven decision-making.

Engaged in understanding and developing machine learning technologies essential for product development. Designed and implemented a data preprocessing service for integrating data from NoSQL and graph databases. Initiated the training of an enterprise-level transformer model for generating text embeddings.

Developed user-level update and question recommendation projects. Utilized Computerized Adaptive Testing (CAT) to recommend personalized sets of questions to users. Applied Item Response Theory (IRT) and Natural Language Processing (NLP) models to estimate relevant parameters for recommendations. Updated user proficiency levels using a mathematical model based on their performance. Enhanced user engagement and accuracy of question difficulty prediction.