Loading organizations...
Loading organizations...
Cactus: Develops a low-latency AI engine and open-source framework for fast, private, cost-effective on-device AI inference for mobile and edge devices.
Based in San Francisco, California, Cactus develops an open-source, low-latency artificial intelligence engine specifically designed for running complex AI models directly on mobile smartphones and wearable devices. The organization provides a specialized software framework that enables fast, private, and highly cost-effective on-device artificial intelligence inference for mobile application developers and enterprise edge computing deployments across various hardware platforms. Operating with a dedicated team of eight employees, the startup recently open-sourced its core inference framework on GitHub to assist thousands of software developers in building localized applications. The firm was selected to participate in the Y Combinator Summer 2025 accelerator batch under the direct guidance of primary partner Andrew Miklas, leveraging technical expertise drawn from previous roles at major technology corporations like Nvidia. Cactus was officially founded in 2025 by Henry Ndubuaku and former quantitative economist Roman Shemet.