From WEKA’s Liran Zvibel on NeuralMesh Axon | RAISE Summit Paris #RAISESummit #Paris #WEKA · · WEKA
“With Axon, we're actually embedding Weka on the GPU servers, eliminating the need for extra networking or specific hardware for storage, which results in higher performance and cost efficiency.”
On , Liran Zvibel, CEO & Co-Founder at Weka, spoke about AI infrastructure during WEKA’s Liran Zvibel on NeuralMesh Axon | RAISE Summit Paris #RAISESummit #Paris #WEKA on WEKA.
Liran Zvibel, co-founder and CEO of Weka, has been speaking about the company's technology and the economics of AI infrastructure. In October 2025, he discussed NeuralMesh Axon, a product that embeds Weka's software directly on GPU servers, which he said eliminates the need for extra networking or dedicated storage hardware and results in higher performance and cost efficiency for large-scale AI. In a September 2025 keynote, Zvibel described 2025 as "the dawn of artificial general intelligence" and the beginning of a fourth industrial revolution, emphasizing that the cost of generating tokens—a concept he termed "tokconomics"—is critical for sustainable AI. He stated that Weka's technology, including a "token warehouse" that caches previously calculated tokens, can reduce GPU cycles and power consumption. Zvibel has also discussed the shift in AI adoption from specialized technology makers to a broader enterprise audience, noting that companies must build future-proof infrastructure to avoid costly overhauls. He highlighted Weka's role in enabling hybrid cloud deployments and improving GPU utilization, and said the company is seeing projects at exabyte scale. In a joint discussion with Micron, he stated that a petabyte of Weka's system saves approximately 260 kilotons of carbon dioxide emissions per year by improving data center efficiency.