Tharun Sai

Built a predictive model to detect fraudulent banking transactions using Python-based data analysis. Performed preprocessing, feature analysis, and model evaluation with metric-driven experimentation. Used NumPy and Pandas for data handling and scikit-learn for modeling and evaluation. Aimed to improve prediction performance through iterative feature and evaluation steps.

Performed comparative analysis of multiple deep learning architectures for CUB 200 image classification. Evaluated model performance metrics to identify better-performing approaches. Documented findings with an emphasis on model comparison and metric-based evaluation. Used the results to guide improvements in experimentation.

Contributed to a messaging feature that integrated a SaaS messaging service. Implemented UI capabilities and connected them to backend APIs for managing messaging workflows. Worked with React.js and Node.js, using SQL and Swagger-documented endpoints for configuration and service integration. Focused on reliable end-to-end messaging flow between the UI and APIs.

Developed scalable ETL/ELT pipelines for insurance-related business data processing. Implemented Snowflake SQL transformations and stored procedures, and optimized query execution. Automated file processing and added data quality validation checks. Integrated Azure Blob Storage and StreamSets for ingestion and workflow orchestration.

Built end-to-end batch ETL/ELT workflows to generate recurring CSV and Excel reports. Implemented Airflow DAGs for scheduling, orchestration, dependency management, and monitoring. Used metadata/config tables to drive SQL transformation logic and manage staging outputs. Included validation and reconciliation steps to ensure reporting data integrity.

Assisted in building SQL/Python-based ETL workflows for ingestion and processing. Developed and tested REST APIs for configuration-driven onboarding and validation, documenting endpoints with Swagger. Supported file-based processing automation and helped ensure pipeline reliability by monitoring jobs and assisting with issue resolution. Collaborated with platform teams to troubleshoot production problems.

Worked on ETL pipeline development for reporting and analytics use cases. Created extraction, transformation, and load workflows that generated periodic CSV/Excel outputs. Added data validation and reconciliation steps such as record-count, minus checks, and schema validation. Monitored pipeline executions and supported operational reliability through logging, alerting, and SLA tracking.

Designed and maintained end-to-end ETL/ELT pipelines for batch data processing and reporting. Built Airflow DAGs for orchestration, scheduling, dependency management, and pipeline monitoring. Implemented staging and metadata-driven transformation logic, and performed validation/reconciliation checks to ensure data quality. Optimized Snowflake SQL performance and automated recurring report generation with error handling and retries.