Compression

ALP Compression

Adaptive Lossless floating-Point compression turns decimal floats into tightly bit-packed integers — often 4–8× smaller than an 8-byte float. Step through the pipeline and watch every transformation.

Metric values like CPU usage (34.5%, 35.1%, 34.8%) look like decimals, but they're actually stored as 64-bit floats — 8 bytes each, most of which carry precision you never needed. ALP exploits the fact that these values are secretly integers in disguise (3450, 3510, 3480) and stores them in a fraction of the space.