Data Scientist
Current• Machine Learning (ML) Expertise: Spearheaded the development of ML models utilizing advanced techniques such as Balanced Bagging Classifiers, Decision Trees, Random Forest, XGBoost, SVM, KNN, KMeans, Grid Search, and Bayesian Optimization. These models have been instrumental in optimizing decision-making processes and significantly reducing incident rates within the organization.• Natural Language Processing (NLP) Projects: Led comprehensive NLP initiatives, including Topic Modeling… Show more • Machine Learning (ML) Expertise: Spearheaded the development of ML models utilizing advanced techniques such as Balanced Bagging Classifiers, Decision Trees, Random Forest, XGBoost, SVM, KNN, KMeans, Grid Search, and Bayesian Optimization. These models have been instrumental in optimizing decision-making processes and significantly reducing incident rates within the organization.• Natural Language Processing (NLP) Projects: Led comprehensive NLP initiatives, including Topic Modeling, Clustering, and Trend Analysis, which have uncovered critical insights from unstructured text data. These projects have driven a 20% increase in customer satisfaction and a 15% reduction in complaints by targeting key pain points and emerging trends.• AI Chat Bot Policy and LLM Implementation: Led the successful organization-wide implementation of AI Chat Bot policies and LLM technologies, guiding the selection of tools and ensuring data security while enhancing operational efficiency.Key Projects:• 811 OneCall Ticket Risk Prediction: Developed a predictive model that identified 15% of tickets responsible for over 70% of damages, enabling proactive risk management.• Voice of Customer (VoC) Customer Segmentation & Topic Modeling: Applied clustering and topic modeling to segment customer feedback, resulting in personalized improvements and a 20% boost in satisfaction.• NLP-Powered Help Desk Ticket Pattern Detection: Implemented NLP and trend analysis on Help Desk data, reducing widespread issues by 40% through early detection and intervention.• LLM-Powered Data Cleansing: Utilized LLMs to cleanse and categorize free-form text data, enhancing data accuracy and improving the reliability of downstream analytics by 30%.• Budget Forecasting: Leveraged data-driven techniques to forecast budget requirements, ensuring alignment with organizational goals and resource allocation. Achieved a forecast accuracy within 10% of the actual annual budget. Show less
