Sriman Reddy

Architected and deployed end-to-end predictive modeling pipelines using Python and XGBoost, improving forecast accuracy by 10% and accelerating analytical throughput. Engineered a financial analytics chatbot using NLP and transformer-based models (e.g., LLaMA 2.0), enabling distillation of key insights from complex financial news articles and reports. Enhanced chatbot capabilities with retrieval-augmented generation (RAG), document summarization, and topic modeling. Applied deep learning and transformer architectures to text classification for financial risk assessment, boosting precision and automating critical compliance validation checks. Developed GAN-based synthetic data for robust model training. Created real-time Power BI dashboards connected to AWS Redshift via custom Python ETL scripts, enabling stakeholders to monitor KPIs with high fidelity. Managed the full data science workflow from data preprocessing and feature engineering to model training and cloud deployment.

Developed predictive models using Python and Spark, achieving an average of 18% improvement in forecasting accuracy. Automated data capture, storage, and validation to enhance data quality by 90% and speed preprocessing. Optimized database architecture and SQL queries, increasing retrieval speed by 50%. Built interactive analysis tools with Streamlit and Tableau, reducing data interpretation time by 80%. Implemented transformer-based models and GANs for text classification and synthetic data generation. Led deployment on Google Cloud Platform, cutting deployment time by 25% and enabling real-time analytics. Led anomaly detection and data augmentation initiatives to improve robustness.