Projects

Developed a Kafka-based real-time data processing pipeline that efficiently consumes, processes, and publishes messages while ensuring robust error handling and summary statistics.

Implemented modular components including a consumer, producer, transformer, and summary printer, utilizing multiple Kafka topics for effective data flow management.

Designed the pipeline to anonymize sensitive data through hashing, ensuring compliance with privacy standards while providing real-time insights into processing metrics.

Developed an automated cold email system using Google Sheets API to manage recipient data and track outreach status, streamlining the job application process. 

Implemented the automation script in GitHub Codespaces, leveraging GitHub Actions for scheduled email sending, ensuring consistent and timely outreach. 

Created a full-cycle automation that sends personalized emails with attachments and updates the Google Sheet with real-time status, improving efficiency and organization of job search efforts.

End-to-End MLOps Pipeline: The project implements a complete machine learning pipeline for predicting coupon acceptance in vehicles, covering data preprocessing, model training with XGBoost, and deployment, while adhering to MLOps best practices.

Robust Technology Stack: It utilizes a comprehensive tech stack including Python, MLflow for experiment tracking, Prefect for workflow orchestration, Docker for containerization, and AWS services for cloud infrastructure, ensuring scalability and reproducibility.

Automated CI/CD and Monitoring: The pipeline features a well-defined development flow with automated CI/CD processes using GitHub Actions, and incorporates continuous monitoring and model retraining capabilities to handle data drift and maintain model performance.