DATA PROCESSING REIMAGINED

Pure Node.js · Zero Dependencies · Columnar Engine

High-performance columnar data engine for Node.js. SharedArrayBuffer + Worker Threads. Built for massive datasets.

$ npm i @cervid/data
5.02s
Load 7.8M rows
0
Dependencies
MIT
License
Performance

BENCHMARK
7,832,546 ROWS

ENGINE RELATIVE TIME TOTAL
@cervid/data
5.02s
Polars (Py)
5.95s
Pandas (Py)
25.80s

CSV · 7.8M rows · Feature Engineering + Filtering. Cervid uses SharedArrayBuffer + Worker Threads.

> CERVID_PROJECT_MANIFESTO
├─ @cervid/decomposer ← Binary format deconstruction
└─ @cervid/vision ← Image processing and computer vision