Alon Konchitsky PhD
Chief Technology Officer at Stealth AI Startup
San Francisco Bay Area, United States
About Alon Konchitsky PhD
Alon Konchitsky PhD is Chief Technology Officer at Stealth AI Startup. Location: San Francisco Bay Area, United States.
My journey in AI infrastructure started long before AI became the center of the world. In the late 1990s, I designed a GPU system built from discrete components. It was the size of a refrigerator and supported 4K-resolution workloads long before 4K became mainstream. That experience shaped the way I think about computing: performance, architecture, efficiency, and scale have always mattered. Since then, I have worked across HPC chips, GPU systems, voice AI, large-scale data centers, and AI acceleration. In 2015, AI became my next frontier. At SRI, I worked on end-to-end voice recognition technology, contributing to systems that helped define the future of intelligent assistants. My work later expanded into early generative AI, optimized AI data-center architecture, and the infrastructure required to run frontier-scale models efficiently. Today, my focus is on building highly cost-effective infrastructure for running LLMs on GPUs. I believe the next major AI breakthrough will not come only from larger models. It will come from making those models economically sustainable: faster inference, smarter memory management, higher throughput, better scheduling, and lower cost per token. That is the mission I am focused on now: improving the economics of GenAI so LLM inference can scale profitably. My goal is simple: accelerate every layer of the LLM stack and extract the maximum performance from the hardware, so token economics finally become positive at scale.
Skills
Additional experience
AI Infra LLM Deployment
Open AI Network Affiliated
September 2025 – March 2026
Education
- Machine Learning, Stanford University
- Executive Program in Business Administration, University of California, Berkeley, Haas School of Business