July 1 2026, 05:35

Syntiant announced a collaboration with Vibe, a provider of contextual AI workspace solutions, to explore and develop next-generation voice-enabled products designed for professional productivity in real-world environments. By combining Syntiant’s Neural Decision Processors and high-performance MEMS sensors with Vibe’s AI product platform and machine learning models, the companies are working to advance physical AI systems built for continuous on-device operation while connecting to cloud-based large language model AI services when needed.

Syntiant is seeing increased potential to combine its Neural Decision Processors (NDPs) and MEMS microphones, sensors, and embedded and ML models into low-power physical AI solutions that truly translate into higher productivity solutions in the workspace. The collaboration between Syntiant and Vibe is focused on delivering responsive, voice-driven interactions that support tasks such as capturing, organizing, and retrieving knowledge throughout the workday, with an emphasis on low-power operation and reduced reliance on constant cloud connectivity.

Based in Bellevue, Washington, Vibe is an AI collaboration platform that helps organizations capture, remember, and act on knowledge across meetings, projects, and teams. Through its portfolio of AI-powered meeting assistants, collaborative workspaces, and interactive smart whiteboards, Vibe transforms conversations into searchable organizational intelligence, helping teams collaborate more effectively and make better decisions.

“The opportunity is to make AI continuously useful without getting in the way of how people work," says Charles Yang, CEO of Vibe. "By bringing intelligence closer to the device, we can reduce friction in everyday workflows and enable more seamless interaction with AI tools across meetings and tasks. Users can engage naturally through voice, allowing them to capture, retrieve, and act on information without interrupting their flow. Our collaboration with Syntiant is focused on making these experiences more responsive and practical in real-world settings.”

As generative AI becomes more widely embedded in enterprise tools, there is increasing interest in architectures that combine edge processing with cloud-based large language models. At the same time, growth in voice interfaces is driving demand for systems that can operate in always-on or near-continuous modes while maintaining efficiency and privacy.

“You need systems designed for continuous sensing and efficient local processing to make AI useful in real-world environments,” adds Kurt Busch, CEO of Syntiant. “Our Neural Decision Processors enable low-power, on-device inference, allowing systems to respond quickly to relevant events while selectively engaging higher-performance compute when needed. Our collaboration with Vibe demonstrates how physical AI and cloud AI can work together to enable more natural, voice-driven experiences.”

Syntiant's NDPs are ultra-low-power processors designed to run neural networks in embedded battery-powered physical AI systems. Purpose-built to run edge AI workloads, NDPs are optimized for highly parallel, compute-intensive workloads while traditional MCUs and CPUs handle control and sequential processing tasks. The portfolio enables continuous neural inference on streaming audio, vision, and sensor data at ultra-low power, allowing devices to sense, decide, and act autonomously in real time.

Founded in 2017 and headquartered in Irvine, California, Syntiant’s physical AI solutions now also include a broad portfolio of MEMS sensors, including SiSonic MEMS microphones and V2S vibration sensors, along with smart audio technologies for IoT, consumer, automotive, and industrial and commercial applications. These sensing technologies enable various always-on use cases such as voice activation, noise cancellation, speech recognition, and sound separation and enhancement. Combined with embedded intelligence at the edge, Syntiant’s platforms deliver contextual audio and sensor processing that enables responsive, privacy-aware experiences across real-world environments while minimizing power consumption.

With more than 100 million purpose-built Neural Decision Processors and machine learning models deployed, along with billions of MEMS microphones and sensors, Syntiant’s technologies and solutions enable the spreading of real-time, AI-powered capabilities, from earbuds to satellites, across a broad range of consumer and industrial applications.
www.vibe.us
www.syntiant.com

About Joao Martins

Since 2013, Joao Martins leads audioXpress as editor-in-chief of the US-based magazine and website, the leading audio electronics, audio product development and design publication, working also as international editor for Voice Coil,