From the TinyML Era to Next-Generation Language Models: how models and hardware are redefining Edge AI

About this session

Edge AI is undergoing a profound transformation, shifting from solutions based on ultra-lightweight models typical of TinyML to increasingly sophisticated systems capable of running advanced language models directly on local devices. This talk explores the technological evolution that has enabled this transition, examining the impact of innovation on both the software side—with more powerful, efficient, and specialized models—and the hardware side, with the rise of neural accelerators, advanced microcontrollers, and new system-on-chip architectures. Real-world use cases will be presented, highlighting the opportunities emerging for low-power, low-latency intelligent applications in fields such as industry, healthcare, and IoT.

Speakers

Massimiliano Torregiani

Founder & Chief Operating Officer, KALPA