Sensors and AI as the Next Generation Interface: Turning the Physical World Into Conversations
ABSTRACT
For decades, we treated sensors as parts and AI as software. That separation is breaking down. The next-generation interface is not a screen. It is the physical world, captured by sensors, interpreted by models, and delivered through natural interactions like voice. In this talk, I will share a practical view of how edge intelligence is changing product design. When microphones, inertial sensors, and environmental sensors become the input layer, and efficient neural processing becomes the interpretation layer, products can respond
instantly, preserve privacy, and operate within tight power and cost limits. That shift is driving demand for complete solutions that combine sensor hardware, purpose-built edge
AI silicon, and production-ready models.
Using the Syntiant story as a real-world case study, I will walk through what it takes to move from a promising demo to a product that ships: robust sensing in noisy environments, on-device decision making, model deployment at scale, and a supply chain that can support high volume markets. The goal is simple. Build interfaces that feel effortless to the user, and are still practical for the engineer.
Audience takeways
- Why sensors plus models are becoming the default interface layer for many products
- The three hardest constraints, power, noise, and privacy, and how to design for them
- What a complete path to production looks like, from sensing to silicon to models
BIOGRAPHY
Kurt Busch is CEO and co-founder of Syntiant, a company that builds ultra-low power edge AI processors, sensors, and machine learning models that enable always on intelligence in
real products. He has spent more than 30 years building and scaling technology businesses, with deep experience spanning product design, engineering, manufacturing,
and fundraising. At Syntiant, he focuses on turning advanced edge AI into systems that customers can ship at high volume, especially in use cases where power, noise, latency,
and privacy constraints are non-negotiable.