Custom MapReduce Engine

Overview

Built entirely without Hadoop or Spark, this engine implements the core MapReduce paradigm in Python — including shuffle, sort, and reduce phases — and benchmarks it against MongoDB aggregation pipelines on real datasets.