Siri M
Development
United States
Skills
Data Engineering
About
Siri Chandana Marepally's skills align with Programmers (Information and Communication Technology). Siri also has skills associated with System Developers and Analysts (Information and Communication Technology). Siri Chandana Marepally appears to be a low-to-mid level candidate, with 3 years of experience.
View more
Work Experience
Data Engineer Assistant
George Mason University
August 2023 - Present
- * Preprocessed the heart disease data using Python, Pandas, and NumPy, to ensure data integrity by removing missing values and outliers. * Employing Python, Matplotlib, Seaborn, Tableau for rigorous EDA yielding 35 visualizations, reducing data preprocessing time by 25%. * Implementing the k-nearest neighbor (k-NN) algorithm to enhance predictive modeling accuracy beyond 76%. * Managed, and deployed 45 production data workflows with Apache Airflow and GitHub on AWS, ensuring smooth operations and achieving a 20% improvement in workflow efficiency. Technologies: Python, Pandas, Numpy, Matplotlib, Seaborn, Tableau, k-NN, Apache Airflow, GitHub, AWS
Senior Data Engineer
Capgemini
July 2021 - August 2022
- * Developed a vintage car insurance analytics application. Utilized PySpark, Python, and SQL to process 1.8 billion historical insurance records and achieved accuracy rate of 98.5%. * Implemented APIs using Python and Flask to seamlessly expose the underwriting model, facilitating a 43% reduction in response time * Incorporated Postman for API testing, executing over 500 HTTP GET test cases and achieved about 65% accuracy * Utilized Git for implementing a collaborative workflow to manage 15 feature branches concurrently, led to 30% reduction in code conflicts. * Integrated Tableau with AWS, optimizing database connections and embedding dynamic visualizations directly into the web application for user enhancement. Technologies: Python, SQL, PySpark, Flask, Postman, Tableau, Git, JIRA, AWS
India -Data Engineer
Capgemini
January 2021 - July 2021
- * Developed an Associate Management system, using Python and Pandas for data ingestion from HR databases, and attendance systems. * Utilized TensorFlow for predictive analytics models in employee retention and performance forecasting, achieving a 92% accuracy for strategic decision-making. * Integrated PowerBI with SQL for real-time reporting, creating interactive HR dashboards with a 30% faster data retrieval speed. * Built RESTful APIs with Python and Flask, testing 150 endpoints with SOAPUI for a 76% success rate in issue identification and resolution. * Designed Apache Airflow DAGs for automated and streamlined ETL processes Technologies: Python, Pandas, Tensorflow, Flask, PowerBI, SQL, Apache Airflow, SOAPUI, Azure Boards, Visual Studio Code
India -Software Engineer
Capgemini
September 2020 - January 2021
- * Developed an Annual Budget Calculation and Evaluation system using Angular and JavaScript interface, resulting in a 25% reduction in user navigation time. * Developed efficient Python APIs using Django, with a 98% efficiency rate in data processing and response times. * Conducted API testing with POSTMAN, resolving data inconsistencies, authentication failures resulting in a 65% increase in overall testing efficiency. * Implemented Azure Boards for agile project management for user stories, tasks, and sprints Technologies: Python, Angular, Javascript, Flask, Django, POSTMAN, Azure Boards Academic Projects: Leaf Disease Identification Through Computer Vision and Machine Learning(Neural Network) * Used OpenCV for resizing (20% reduction), normalization, and noise reduction, achieving a remarkable 95% improvement in image quality. * Constructed a CNN model using the Keras framework, integrating 15 layers achieving a classification accuracy of 92% in identifying various leaf diseases. * Utilized Python with NumPy, pandas, and SciPy for HSV-based image visualization, realizing a 30% reduction in data processing time for efficiency. Hand Gesture Recognition with Computer Vision * Developed a Support Vector Machine (SVM) in Python, achieving 95% accuracy in the precise classification of hand gestures. * Used OpenCV to identify 50 interesting points and extract contours from hand images, enhancing model training by 20% in gesture recognition.