Motivation#

Many backend systems need to answer one question extremely fast: have I seen this before? Traditional data structures often trade memory for speed, which becomes expensive at scale.

This project was created to explore probabilistic data structures in Go while gaining hands-on experience with bitwise operations, hashing strategies, and memory-efficient design.

Results#

The implementation uses a compact []uint64 bit array with auto-tuned parameters to balance false positives and memory usage. Hashing is inspired by MurmurHash to ensure even distribution and consistent performance.

The result is a practical Bloom filter suitable for caching layers, deduplication systems, and learning-focused explorations of Go internals.