BIG DATA / CLOUD

NYC Taxi Analytics on AWS.

Scalable trip-data transformation and predictive modeling with PySpark on AWS EMR.

AWS EMRPySparkS3Machine Learning

The challenge

Large taxi-trip datasets need distributed preprocessing, consistent feature engineering, and repeatable evaluation.

The approach

Built a PySpark workflow using taxi trip features, categorical encoding, and machine-learning pipelines to predict fare-related outcomes.

Key learning

Cloud analytics requires attention to permissions, data schemas, feature cardinality, and reliable training pipelines.