I built a fuzzy matching engine that's 300x faster than RapidFuzz on 1M records

Fuzzy matching is one of those tasks that feels "easy" until you hit real-world data volumes. If...