# doubleAI launches Maxwell, citing a 67% cut in one customer's compute costs > Source: > Published: 2026-10-19 15:46:44+00:00 # doubleAI launches Maxwell, citing a 67% cut in one customer's compute costs **Co-founders Amnon Shashua and Shai Shalev-Shwartz say the system can improve software as it runs. DoubleAI reported a 67% cost cut for one early customer; LTX told the Journal Maxwell doubled its model-inference speed.** By [Ryan Merket](https://runtimewire.com/author/ryan-merket) · Published Primary source: [The Wall Street Journal](https://www.wsj.com/cio-journal/nvidia-backed-startup-doubleai-launches-model-to-pare-cost-of-compute-b780f215) ## Why it matters DoubleAI is betting that software optimization backed by formal code verification can become a paid alternative to simply buying more compute. Maxwell's results will need to hold across customers' workloads, with a clear way to measure savings and fees. Amnon Shashua and Shai Shalev-Shwartz's [doubleAI](https://www.doubleai.com/?ref=runtimewire) introduced Maxwell on October 19th, according to [The Wall Street Journal](https://www.wsj.com/cio-journal/nvidia-backed-startup-doubleai-launches-model-to-pare-cost-of-compute-b780f215?ref=runtimewire). The system is what its founders call a "code execution model": AI designed to optimize software while it runs and reduce the computing resources it consumes. DoubleAI said Maxwell cut one early customer's compute costs by 67%.… When you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.