This repository contains the analytics projects developed during my internship at Peacock (NBCUniversal). The projects focus on building robust analytics tools for A/B testing and experimentation, leveraging advanced statistical methods and machine learning approaches.
- MDE (Minimum Detectable Effect) Dashboard A comprehensive analytics dashboard built to monitor and analyze cross-device watch time patterns and experiment metrics. Key Features:
- Real-time metrics monitoring across cross-device experiments
- Statistical inference analysis with a 30-day rolling window
- 5 automated metric calculations for experiment tracking
- 3 customizable filters for detailed data exploration
- Daily automated data refresh pipeline
- Automated Inference Engine An automated system for processing and analyzing A/B test results using advanced statistical frameworks. Key Features:
- Automated statistical inference across multiple test scenarios
- Machine learning models for pattern detection
- Real-time data processing pipeline
- Comprehensive test result analysis
Google Cloud Platform (GCP) Python (scikit-learn, pandas, numpy) Statistical testing frameworks ETL pipelines