In the Future of Voice AI series of interviews, I ask three questions to my guests:
- What problems do you currently see in Enterprise Voice AI?
- How does your company solve these problems?
- What solutions do you envision in the next 5 years?This episode’s guest is Dani Cherkassky, CEO and Co-Founder of Kardome.
Dani Cherkassky has spearheaded the development of groundbreaking voice technology designed to excel in complex, real-world environments. Before founding Kardome, Dani led Silentium’s technology strategy as CTO and VP of R&D, introducing the Silentium Quiet Bubble Technology to reduce road noise in car cabins.
His career reflects a deep commitment to transforming how we interact with technology. He combines strategic leadership with technical expertise to deliver impactful solutions.
Kardome is transforming voice technology with its innovative Spatial Hearing Voice AI solution. This advanced software is designed to enhance voice-enabled devices and automotive systems, offering exceptional accuracy in locating, identifying, and capturing individual voices in real-time, even in multi-speaker, noisy environments.
Recap Video
Takeaways
Kardome was founded by engineers frustrated with the poor performance of devices like Alexa and Google Home.
Dani realized voice systems weren't just missing the words—they were missing the world around them.
Even with AI advances, voice recognition systems struggle in noisy or real-world settings.
When state-of-the-art voice recognition systems are taken out of their comfort zone and put into an acoustically challenging setting—like a noisy office or a car—accuracy drops dramatically.
Kardome solves 3D audio challenges, helping systems understand environments and isolate voices.
They’re building human-like AI tech that sees the world through sound, even in the toughest conditions.
Kardome’s technology mimics human hearing by analyzing sound from all directions and making sense of it.
Their system runs on-device for faster processing without the internet.
On-device architecture is challenging to build, but brings major benefits in speed and privacy.
Kardome’s main users are in vehicles and consumer electronics like smart speakers and TVs.
The technology lowers hardware costs by using fewer microphone arrays with advanced software.
Their approach integrates components like voice IDs, wake words, and audio analysis for better efficiency.
Today’s AI is reactive and lacks awareness of its surroundings, limiting its usefulness.
The future of AI agents needs to be proactive, offering help without waiting for commands.
Consumer electronics need better tech to work seamlessly with AI like ChatGPT.
Current voice systems are fragmented; Kardome focuses on more unified solutions.
Kardome’s goal is to solve real-world problems and deliver practical results.
Share this post