Sanjana Singh

Mobile Device Fraud Identification: Developed and deployed a fraud model automating order release and cancellation, resulting in a 30% reduction in fraud. IoT Data Analytics: Utilized AWS services including Glue, S3, Lambda, Athena, QuickSight, and SNS to analyze IoT device data. PDF Data Extraction: Automated manual PDF data extraction using Langchain, PyPDF, and OpenAI Embeddings. Flow Forecast: Developed a net flow prediction model using ML/DL techniques, achieving 70% higher accuracy. Uplift Modeling: Developed a model using Python, Sagemaker, and SQL to predict CATE scores from advertisement features.

Emerging Topic Extraction: Employed NLP techniques to automatically extract emerging topics from survey comments. Customer Complaints Analysis: Investigated valuable insights by categorizing complaints data, enhancing understanding of customer sentiment.

Led initiatives for insurance clients, driving data-driven insights and strategic recommendations. Risk Finance Optimization: Worked in a risk modeling team for a US-based fortune 500 insurance brokerage firm. Performed Mean loss calculations and percentile analysis. Calculated layer values by risk transfer layers. Modelled and validated High Severity Low-Frequency events.

Automated data extraction from PDF reports and PowerPoint presentation creation. Developed VBA tool to extract and analyze data from Loss Projection Modeling reports. Automated tracker updating from Outlook mail, reducing processing time from 2-3 hours daily to 10 minutes.