Priyankka M

About Priyankka M

Software Engineer with 4+ years of experience building scalable applications, backend…

Experience

1. Data Scientist

   Insight Global

   Jan 2025 — Present

   Analyzed claims and pharmacy data using Python and PostgreSQL, identifying risk patient cohorts and reducing hospital readmissions by 12% through predictive modeling and hypothesis testing, while ensuring strict HIPAA compliance in handling sensitive patient data.• Developed ensemble models (Random Forest, XGBoost) with Scikit-learn and TensorFlow to detect fraudulent claims, improving detection accuracy by 18% while reducing false positives that previously delayed genuine reimbursements.• Applied Natural Language Processing (NER) to extract clinical terms from unstructured physician notes, enhancing patient risk stratification models and supporting case managers in delivering personalized care interventions.• Designed and deployed CNN-based models on AWS for automated medical image classification, accelerating prior-authorization approvals by 20% and reducing manual review time for radiology staff.• Integrated LangChain and LLM-powered Gen AI solutions to streamline member support workflows, enabling faster query resolution for benefit eligibility and medication coverage, cutting average handling time by 25%.• Built interactive dashboards in Tableau and Looker to track medication adherence, care gaps, and claims trends, empowering business leaders with real-time visibility into key health outcomes and financial performance.• Performed advanced statistical analysis, including hypothesis testing, to evaluate the impact of intervention programs, validating a 9% improvement in medication adherence across targeted patient groups.