I built NovaEngine to solve the high cost of cloud log storage. It achieves ~40 MB/s throughput using a pattern recognition approach rather than standard compression. In my tests, I've reached up to 90% space savings on redundant datasets. Looking for technical feedback on the architecture and performance!