SRUJAN CHINTA

Research Assistant at University of Texas at Arlington

Built a conversational AI system using LangChain and LLaMA 3 to automate facility location and shipping cost optimization. Collected and transformed supply chain data into modeling-ready datasets, ensuring quality and consistency. Designed NLP and regex pipelines to extract structured variables for predictive modeling and optimization tasks. Applied regression-based models and optimization techniques to evaluate routing efficiency and cost tradeoffs. Communicated findings through reports and planned dashboards, translating results for technical and non-technical audiences.

Data Scientist at Ceyline Shipping Services Pvt. Ltd.

Led EDA and statistical analysis on 10K+ daily shipment records; uncovered seasonal patterns and process gaps, reducing late deliveries by 10%. Engineered production-grade preprocessing (outlier handling, imputation, scaling) to raise model data quality and consistency. Partnered with product and operations to define KPIs and design experiments (A/B, pre-post) for process improvement validation. Built and tuned a Random Forest delay prediction model, improving reliability by 15% via feature engineering and GridSearchCV. Delivered a Prophet forecasting pipeline for weekly demand, increasing fleet utilization by 8% and lowering underutilization costs. Shipped Power BI dashboards for KPI monitoring, model outputs, and stakeholder reporting to accelerate decision cycles.