Sai Teja Sankarneni

Experience

• 

  Kpit Technologies

  Jul 2019 - Sept 2020

  Jr. Data Scientist

  o Designed a demand forecasting model using ARIMA, optimizing spare parts inventory and reducing excess stock by 15%. o Created interactive Tableau dashboards for fleet performance tracking, improving logistics efficiency by 20% and streamlining maintenance scheduling. o Conducted comprehensive performance tuning and optimization of SQL Server databases, including indexing, and query restructuring to significantly reduce query execution times and enhance overall system responsiveness. o Effectively translated complex machine learning model outputs, statistical analyses, and predictive insights into clear, actionable recommendations for cross-functional stakeholders, including business executives, product managers, and marketing teams, resulting in data-driven decision-making and a 15% improvement in campaign ROI. o Streamlined the ML model deployment process by integrating AWS CodePipeline, achieving a 50% reduction in deployment time and supporting continuous integration/continuous delivery (CI/CD) for machine learning projects. Show less

• 

  Accenture

  Oct 2020 - Aug 2022

  Data Scientist

  o Defined fraud detection KPIs in collaboration with risk management teams, integrating compliance standards (Basel II, AML, GDPR) into monitoring frameworks. o Created a Customer Lifetime Value (CLV) model to segment high-value customers, integrating business intelligence insights to optimize marketing strategies, enhance customer retention, and increase cross-sell and upsell conversions by 15%. o Refined Accenture’s credit card fraud detection pipeline, reducing false positive alerts by 30% and improving customer trust. o Developed a deep learning-based predictive maintenance model using PyTorch, forecasting equipment failures with 90% accuracy and reducing unplanned downtime by 18%. o Designed a demand forecasting model using ARIMA, optimizing spare parts inventory and reducing excess stock by 15%. o Built an anomaly detection system using Isolation Forest, identifying inefficiencies in-vehicle telematics data and cutting maintenance costs by 22%. Show less

• 

  Infosys

  Apr 2024 - now

  Data Scientist

  o Optimized high-performance SQL queries to process millions of daily transactions, reducing query execution time by 35% and accelerating financial reporting o Developed Tableau dashboards to deliver real-time fraud alerts, revenue insights, and customer churn risk scores, improving decision-making by 40%. o Built a predictive financial model to enhance revenue forecasting accuracy by 15%, improving budgeting and investment strategies. o Designed a credit risk assessment model using Logistic Regression and XGBoost, improving default prediction accuracy by 12% and minimizing bad loan approvals. o Automated ETL pipelines to ingest, clean, and transform financial data from multiple sources, ensuring 99.9% accuracy, improving data consistency, and reducing manual data handling. Show less