Tharun Ponnaganti

Master's

Statistics

American Automobile

Data Scientist

- Used the Agile methodology to build the different phases of the Software development life cycle - Developed a sql_processing function integrating with Vertex AI to automate SQL query generation from natural language requirements - Leveraged the code chat-bison model within Vertex AI for intelligent code and query responses, enhancing efficiency - Engineered sophisticated predictive analytics models to optimize the lead scoring system for P&C insurance campaigns, utilizing Python and SQL for data manipulation and analysis of a dataset comprising 6.2 million records and 97 features - Implemented and tuned logistic regression, random forests, and gradient boosting, achieving an AUC of ROC 0.88 - Incorporated AI-generated responses and language model interactions into a Python-based application using LangChain and OpenAI's LLMs, streamlining secure API integrations to enhance user experience with dynamic, context-aware text processing - Generated 30+ Tableau reports, enhancing executive decision-making with clear visualizations of KPI's