JuliaDB is designed for Julia, in Julia. It works with functions and data types from any of the thousands of Julia libraries or custom ones written for your specific use case. Yet, queries run as fast as the hardware affords by utilizing Julia's built-in just-in-time compiler. Fast user-defined-functions makes it suitable even for workloads requiring custom algorithms and machine learning.
JuliaDB has built-in distributed parallelism. You can use Julia with many processes on a multi-core machine or a cluster
with practically no setup. You can interactively prototype on a small sample of the data, and scale to any amount of data without any change to the code, by adding more computational power to the setup. (Plug: JuliaRun
is Julia Computing's scalable deployment product that works well with JuliaDB).
JuliaDB can read CSV files (fast), save and load data using an efficient binary format, extract statistics in parallel, perform regressions, all out-of-the box. It ties together the most useful data manipulation libraries for a comfortable experience.