- Published on
A short history of Parquet, then two experiments on one day of Ethereum logs: how bloom filters speed up lookups on a high-cardinality column, and how sorting and compression shrink a 1.2 GB file to 218 MB.
Data-engineering
All Posts
- recommendation-system (6)
- machine-learning (5)
- x-algorithm (5)
- ai (3)
- vibe-coding (3)
- claude-code (3)
- system-design (3)
- rust (3)
- mindset (1)
- technology (1)
- vietnam (1)
- business (1)
- culture (1)
- graph (1)
- embedding (1)
- pytorch (1)
- llm (1)
- parquet (1)
data-engineering (1)
- blockchain (1)
- big-data (1)
- optimization (1)
- til (1)
- mathematics (1)
- engineering (1)
- geometry (1)
- infrastructure (1)
- commerce (1)
- shopify (1)
- agentic (1)
- strategy (1)
- side-project (1)
- project-management (1)
- bmad (1)
- terraform (1)
- deployment (1)
- aws (1)
- transformer (1)