Sriram Pulipaka

Performed data extraction, transformation, and loading (ETL) using Apache Airflow and SQL, maintaining a 98% data accuracy rate. Developed and maintained SQL scripts for complex data retrieval, enabling 15% faster query performance. Built Power BI dashboards to monitor sales performance, customer segmentation, and revenue metrics, reducing manual reporting time by 40%. Gather data from multiple sources (databases, APIs, flat files) and ensure consistency and accuracy during preprocessing and Analyze data using Python and SQL to identify trends, correlations, and outliers. Conducted root cause analysis (RCA) to identify data discrepancies, resolving data quality issues within SLA timelines. Implemented Python-based automation to streamline data validation and report generation processes.

●Built Python ETL pipelines to parse and normalize logs for Splunk analytics. ● Automated log cleaning and anomaly detection, reducing false positives by 30%. ● Integrated real-time dashboards and alerting for faster incident response. ●Documented workflows for consistent data validation and team alignment. ●Automated performance profiling and optimization, reducing pipeline execution time and lowering infrastructure costs. ●Enhanced log analysis with advanced pattern-matching, enabling precise extraction of security event indicators from diverse log formats. ●Implemented comprehensive logging and exception handling to streamline troubleshooting and maintain robust audit trails. ●Facilitated cross-team training sessions to share domain knowledge, foster collaboration, and onboard new users to the analytics workflow.

●Supported Python log analytics and automation for server, network, and firewall data. ● Built basic dashboards and queries in Splunk to visualize security and operations trends. ● Automated routine data collection and reporting, reducing manual workload for daily analytics. ● Collaborated with analysts to improve data quality, documentation, and cross-team information flow.

To uncover patterns and correlations in violent crime, informing more targeted violence reduction strategies by analyzing how offender age and race relate to domestic violence incidents, victimization rates, and crime severity. The crime-demographic-trends-chicago project used Python for data cleaning, visualization, and correlation analysis; R for statistical testing and additional visualizations; and SQL (with AWS RDS and MySQL) for database management, aggregation, and querying. Additionally, AWS Glue DataBrew was leveraged for data exploration and profiling, while AWS RDS provided scalable cloud database storage. This multi-tool approach enabled comprehensive data handling, from raw data collection to advanced statistical and visual analysis, supporting evidence-based insights into violent crime trends by offender demographics.

The data-analysis-and-ml-workbench is designed to provide a structured, hands-on platform for learning and practicing data science and machine learning skills through real-world examples—focusing on data cleaning, exploration, visualization, model building, and performance evaluation. Core technologies typically include Python as the programming language, Jupyter Notebook for interactive development and documentation, NumPy and Pandas for data manipulation and analysis, Matplotlib and Seaborn for data visualization, and Scikit-learn for implementing and evaluating machine learning models.

This project explores local unemployment trends in California using the LAUS dataset. The analysis includes linear modeling, time series visualization, statistical testing, and categorical analysis. This project primarily utilizes Python (with libraries such as pandas and SciPy) for data cleaning, statistical testing, and modeling; matplotlib and seaborn for time series and categorical visualizations; and Jupyter Notebook as an interactive environment for exploratory and reproducible analysis.