By Craxel Founder and CEO David Enga
March 30, 2025
Algorithmic innovation is proving to be highly disruptive as we begin the journey to a highly automated future. But it is not just algorithms like attention in transformers that are going to power AI-powered decision making. Data is the critical input (data is the new oil) and organizations are struggling today with its size and complexity, and the problems keep getting worse every day. So how are today's struggling data algorithms going to miraculously overcome these challenges in time to support AI-powered decision making at scale?
They are not. We believe these legacy algorithms are obsolete. Most people don't realize that the algorithms that power today's databases were invented in the 1970's. To scale them requires lots and lots of expensive servers. The dirty little secret is that the big powerful multi-core processors in those servers are highly underutilized. Do you know why? It is as simple as this. The algorithms are simply unable to do many things at the same time. No matter how many cores, they have bottlenecks that limit parallelization. In computer science, we actually have a law for this. It is called Amdahl's law. The processing cores are sitting there waiting much of the time. Yet, organizations are still paying for it. This is an incredible waste of money and resources. The alternative to these algorithms has been to brute force scan vast quantities of excess data, adding extreme latency or cost or both. Craxel has built the knowledge infrastructure for AI-powered decision making using an algorithmic innovation akin to a perfect order preserving hash that scales, but that also works in many dimensions. It allows many, many things to be done at the same time in and across computers. The software is called Black Forest and compared to the prevalent horse and buggy data engines in use today, its like a starship with a warp drive. The future that Black Forest is going to deliver is fast, efficient, secure, and resilient AI-powered decision making at any scale. It’s also a strategic imperative that the U.S. government deploys it everywhere as fast as possible.