<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Kardome Technology LTD on Medium]]></title>
        <description><![CDATA[Stories by Kardome Technology LTD on Medium]]></description>
        <link>https://medium.com/@kardome?source=rss-45676c928070------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*03x1lz2Ie1ohcAP26jBCtw.jpeg</url>
            <title>Stories by Kardome Technology LTD on Medium</title>
            <link>https://medium.com/@kardome?source=rss-45676c928070------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 09:30:41 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kardome/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Why Infotainment Is the One Part of the Car That Isn’t Getting Better]]></title>
            <link>https://medium.com/@kardome/why-infotainment-is-the-one-part-of-the-car-that-isnt-getting-better-77da30508ccc?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/77da30508ccc</guid>
            <category><![CDATA[voice-ui]]></category>
            <category><![CDATA[voice-assistant]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[voice-technology]]></category>
            <category><![CDATA[infotainment]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Sat, 03 Oct 2026 13:03:51 GMT</pubDate>
            <atom:updated>2026-10-03T20:10:20.121Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NlvY9TgK6BLHZExHYp0SMA.jpeg" /></figure><p>Automakers have spent the last decade making cars better at almost everything. Fewer powertrain issues, fewer climate complaints, fewer problems with seats and exteriors. Then there’s the dashboard.</p><p>JD Power’s 2026 <a href="https://www.jdpower.com/business/press-releases/2026-u-s-multimedia-quality-and-satisfaction-study/">U.S. Multimedia Quality and Satisfaction Study</a> makes the gap clear. Infotainment accounts for 25% of all reported vehicle problems, and the five most common complaints in the industry all come from it. Multimedia problems ticked up from 42.3 per 100 vehicles in 2025 to 42.4 in 2026, while nearly every other category improved.</p><p>Infotainment isn’t collapsing. It’s stuck. It’s the one system in the car that the industry hasn’t figured out how to improve.</p><h3>The 90-day problem</h3><p>JD Power surveys owners after 90 days, and that timing matters. It’s long enough for the novelty to wear off and the daily friction to set in. It’s the stretch where a voice command that fails once a week stops being a quirk and becomes the thing an owner complains about.</p><p>To be fair, the study doesn’t break out voice on its own. But JD Power points to system complexity and software performance, not hardware, as the source of today’s quality issues. Voice UI is where that complexity meets real people in a real cabin.</p><h3>Why today’s voice UI breaks in a full car</h3><p>Most in-car voice systems are built to hear sound, not people. The wake word triggers, noise suppression cleans up the audio, and the ASR engine turns it into words. What’s missing is who said it and where they’re sitting. So a passenger in the back says, “It’s freezing in here,” and the car turns up the driver’s-side window. A conversation in the back gets picked up as a command. The driver repeats themselves over the radio, gives up, and reaches for the touchscreen.</p><blockquote>“Every OEM team we talk to tells the same story,” said Dani Cherkassky, CEO of Kardome. “Voice works in testing and breaks on the road, because engineers end up overfitting to the test. Kardome replaces that patchwork with one holistic, context-aware stack that understands the occupants, the speakers, and the state of the car, so performance holds up where it actually counts.”</blockquote><h3>Where Kardome fits</h3><p>Kardome Voice AI solves this problem. It gives the car human-level understanding. Anyone in the car can ask for something mid-conversation, and the right thing happens, in the right seat, the first time.</p><p>It’s a full voice AI stack in a single on-device package that fits existing hardware. That lets engineering teams build context-aware, proactive, conversational voice UI with minimal integration effort, faster time to market, and less R&amp;D overhead.</p><h3>Experience it yourself</h3><p>The best way to understand the difference is to hear it. Check out our video demo below. If you work in HMI/voice in the digital cockpit, reach out to our team for an in-person demo <a href="https://www.kardome.com/company/contact-us/">here</a>.</p><p><a href="https://www.kardome.com/resources/videos/engineering-the-future-of-in-cabin-ui-ux-kardome-automotive-voice-ai-demo-2/">Future of In-Cabin UI/UX | Kardome Automotive Voice AI Demo</a></p><p>Infotainment is the one part of the car that stopped getting better. Voice UI is a good place to start changing that.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=77da30508ccc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Shift to Perception-Layer AI: Moving Beyond Traditional ASR Architecture]]></title>
            <link>https://medium.com/@kardome/the-shift-to-perception-layer-ai-moving-beyond-traditional-asr-architecture-8e2484b7b86e?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/8e2484b7b86e</guid>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[physical-ai]]></category>
            <category><![CDATA[spatial-ai]]></category>
            <category><![CDATA[contextul-ai]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Fri, 24 Jul 2026 17:31:51 GMT</pubDate>
            <atom:updated>2026-07-24T17:31:51.948Z</atom:updated>
            <content:encoded><![CDATA[<p><em>The challenge is no longer simply detecting speech. It is understanding the entire contextual scene: a continuous, real-time model of the physical environment and everyone in it.</em></p><figure><img alt="Physical AI requires perception, reasoning, and action" src="https://cdn-images-1.medium.com/max/1024/1*nzCgJz3SGXtsCZQdgjybmw.jpeg" /></figure><p>Contextual intelligence is a term being thrown around from tech labs to boardrooms, but in practice, it represents a departure from how the industry has built voice products for decades.</p><p>For years, the consumer electronics industry has treated voice AI as a speech recognition problem: improving wake words, reducing noise, optimizing ASR, and sending cleaner audio to the cloud. That approach worked reasonably well when voice assistants lived in controlled environments with one user, generic commands, and a quiet room.</p><p>But consumer devices no longer operate in those conditions.</p><p>Today’s smart TVs, soundbars, appliances, and connected home devices exist in acoustically chaotic environments with multiple people speaking simultaneously, media playback, kitchen noise, reflections, open windows, moving users, and overlapping conversations. And this is where traditional voice AI architectures begin to break down. Not because ASR models are weak, but because most systems still lack a contextual understanding of their environment: where sounds are coming from, who is talking, and what the user actually intends.</p><h3>The Missing Layer</h3><p>Humans naturally separate sound sources spatially. In a crowded room, people instinctively know who is speaking, where the voice is coming from, and what it means given the surrounding context. Most voice AI systems still do not process audio this way.</p><p>Conventional pipelines treat sound as a mixed signal that requires enhancement before recognition. Traditional beamforming and noise suppression help at the margins, but they are fundamentally limited in dynamic, multi-speaker environments because they attempt to reconstruct the acoustic scene after discarding the spatial information that defined it.</p><p>The challenge is no longer simply detecting speech. It is understanding the entire contextual scene: a continuous, real-time model of the physical environment and everyone in it.</p><blockquote>Room reflections have been treated as noise to be suppressed rather than as data to be interpreted.”</blockquote><h3>The Architectural Bottleneck</h3><p>For a decade, the industry has been locked in a race to clean up audio: more microphones, refined beamforming, and optimized cloud-based filtering. Yet the performance ceiling remains.</p><p>The problem is a fundamental design assumption: room reflections have been treated as noise to be suppressed rather than as data to be interpreted. When voice devices fail in real-world environments — noisy kitchens, living rooms with media playback, automotive cabins at highway speeds — it is because they are attempting to listen without first perceiving the physical environment.</p><h3>The Shift: Building a Perception Layer</h3><p>A different approach inverts the conventional sequence. Rather than cleaning the signal first, a perception layer analyzes the environment before passing anything to the recognizer, building a spatial model of who is speaking and where by using reflections that conventional pipelines discard.</p><p>Spatial Hearing AI maps the physical scene by analyzing acoustic reflections rather than suppressing them. Each sound source generates a unique acoustic fingerprint, derived from how its signal interacts with the environment, allowing the system to identify who is speaking and precisely where they are in 3D space, whether seat-by-seat in a vehicle cabin or room-by-room in a smart home.</p><h3>The Proof: Reflections Are Data, Not Noise</h3><p><a href="https://www.kardome.com/resources/whitepapers/accurate-speech-recognition-at-120-km-h/">HEAD acoustics </a>tested this spatial hearing architecture in a Renault Megane at speeds up to 120 km/h with four simultaneous speakers, a single existing microphone array, and full ITU-T P.1100 compliance. Standard hands-free systems reached 22.7% speech recognition accuracy under those conditions. The spatial hearing architecture reached 98%.</p><p>The result reflects the core architectural difference. Conventional systems attempt to recover from noise after the fact. A spatial hearing approach uses the acoustic environment as a data source from the outset, making reflections do work rather than filtering them away.</p><h3>Building on Spatial Awareness: Cognitive AI</h3><p>Accurate spatial separation is necessary but not sufficient for production-grade voice AI. Once the system knows who is speaking and where, a second processing stage handles on-device inference: tracking speakers across time, separating voice from background events, identifying users through biometric signatures, and understanding intent in context, without routing data to the cloud.</p><p>Together, these two stages allow a device to hear the way a human does: by understanding the physical environment before attempting to process speech.</p><h3>The On-Device Foundation</h3><p>Historically, contextual understanding has been difficult to achieve through cloud-centric architectures, where voice interactions are processed as discrete requests rather than as part of a continuously evolving acoustic scene. On-device inference changes this by allowing systems to maintain awareness between interactions rather than only responding when a command is detected.</p><p>For OEMs, this creates direct strategic advantages:</p><ul><li><strong>Differentiated user experiences.</strong> Devices can support more natural interactions that adapt to users, environments, and usage patterns rather than requiring controlled conditions.</li><li><strong>Greater reliability in real-world conditions.</strong> Continuous local processing keeps systems responsive in noisy, dynamic, multi-user environments where cloud-dependent architectures degrade.</li><li><strong>Privacy and regulatory alignment.</strong> More processing remains on the device, reducing the need to transmit sensitive user and environmental data.</li><li><strong>Reduced cloud dependency.</strong> Local inference lowers infrastructure costs while reducing latency and connectivity requirements.</li></ul><h3>Use Case: The Modern Smart TV</h3><p>The transition to contextual intelligence is not theoretical. Consider the modern smart TV, a device that operates in one of the most acoustically challenging environments in the home.</p><p>Unlike traditional voice assistants designed for isolated interactions, a TV exists in a shared living space where multiple people may be speaking while media plays at high volume.</p><p>Recognizing speech is only part of the challenge. To deliver a reliable voice experience, the device must understand context. It must distinguish between television audio and human speech, determine when a user is directing a request toward it, and maintain reliable interactions despite competing sound sources and changing conditions.</p><p>LG has integrated <a href="https://www.kardome.com/technologies/spatial-hearing/">spatial hearing technology</a> into select <a href="https://www.forbes.com/sites/scottkramer/2026/06/14/first-look-spatial-hearing-on-a-tv/">OLED TV</a> models to address this problem, enabling reliable voice interactions in real-world home environments where conventional voice systems struggle with interference from media playback.</p><p>Achieving this level of contextual understanding requires technologies such as:</p><ul><li>speaker separation,</li><li>localization,</li><li>real-time acoustic scene analysis.</li></ul><p>These capabilities work together to help devices understand not only what was said, but the context in which the interaction occurred.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EKcsoJQU8-ukYH5rcepi7w.png" /><figcaption><em>Key Use Cases for Next Generation Voice AI</em></figcaption></figure><h3>Voice AI Is Entering a New Phase</h3><p>The first generation of voice-enabled devices focused on recognition and rigid commands. The next generation will require a continuous, contextual understanding of the physical environment and the people in it.</p><p>That transition is already reshaping consumer electronics design, particularly in devices operating in shared, noisy, and multi-user spaces. As manufacturers adopt voice AI as a foundational interface layer across product categories, the differentiator will not be the size of the language model or the number of wake words supported. It will be whether the device can reliably understand people in the environments where they actually live.</p><p><em>Authored by </em><a href="https://www.linkedin.com/in/khamharn/"><em>Oddy Khamharn, Kardome VP | Product Strategy and Partnerships</em></a></p><p>How is your organization rethinking its voice interface? We’d love to hear your thoughts on the future of contextual AI. <a href="https://www.kardome.com/company/contact-us/">Send us a Message</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8e2484b7b86e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Voice AI Value Equation]]></title>
            <link>https://medium.com/@kardome/the-voice-ai-value-equation-37960d07ea29?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/37960d07ea29</guid>
            <category><![CDATA[voice-technology]]></category>
            <category><![CDATA[voice-tech]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[voice-recognition]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Fri, 15 May 2026 05:04:45 GMT</pubDate>
            <atom:updated>2026-05-15T05:04:45.136Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*74vLNpblpkm_jhn4nWkPig.png" /><figcaption>In this month’s issue: the economics of in-car voice AI, five takeaways from the SBD Automotive profitability panel, three signals we’re tracking across the industry, and where to find us at InnoVEX.</figcaption></figure><h3>The Profitability Path of Automotive AI: Webinar takeaways</h3><p>Most OEM voice assistants have become “Zombie” features: capability without trust, cost without revenue. Drivers have already voted with their behavior. They route navigation through Siri and Google Assistant, not the native head unit.</p><p>Voice AI is a critical lever in that profitability equation, but only if it actually works. Here are five takeaways from Kardome CEO Dani Cherkassky’s contributions at <a href="https://www.linkedin.com/company/sbdautomotive/">SBD Automotive</a>’s April 15 panel worth carrying into your roadmap:</p><ol><li>Capability isn’t revenue. Revenue depends on delivering a genuine improvement to the user experience.</li><li>Cloud-first LLMs are too slow. Their 1–3-second latency erodes user trust by exceeding human conversational pause times (200–300ms).</li><li>Wrong architecture for the job. Using heavyweight cloud LLMs for simple actions (like adjusting temperature) is inefficient, slow, and costly.</li><li>Deaf and blind when asleep. Cloud-dependent AI must sleep to manage cost/compute, losing continuous cabin awareness and context.</li><li>The path forward is hybrid. A “System 1” SLM (small language model) at the edge handles fast, intuitive interactions, while a “System 2” cloud LLM handles complex reasoning, improving cost and latency.</li></ol><p>Watch Dani’s take:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fx-Er4oabfNw%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dx-Er4oabfNw&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fx-Er4oabfNw%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/1e7ef5efb4ed78ba9c4920e74aebe9c8/href">https://medium.com/media/1e7ef5efb4ed78ba9c4920e74aebe9c8/href</a></iframe><p>For a deeper dive into why the architecture itself is the problem, and what it takes to flip in-car voice from cost center to profit center, read Dani’s full analysis: → <a href="https://hubs.ly/Q04f-N9q0">The Voice AI Monetization Gap — Usability and Engagement are Key to Profitability</a></p><h3>Voice AI Trends to Watch</h3><p>Three signals from across the automotive Voice AI space that we think are worth tracking this month:</p><ul><li>Edge AI as a differentiator. McKinsey’s recent <a href="https://www.mckinsey.com/industries/semiconductors/our-insights/the-rise-of-edge-ai-in-automotive?utm_campaign=40465778-Kardome%20Voice%20AI%20Authority%20Drive&amp;utm_source=hs_email&amp;utm_medium=email&amp;_hsenc=p2ANqtz-8Tadr-SdEZ_CfIrDKrtKFoiCJ4flK5e-jNbaHhOGHlfMhCGwXB6U4ffDFF4yhoqeOWG2tS">edge AI in automotive brief</a> reframes on-device inference as a critical cost lever. OEMs are increasingly prioritizing on-device benchmarks alongside accuracy to manage bill-of-materials costs.</li><li>The agentic shift. Intent-driven, multi-step flows are moving into production. <a href="https://www.abiresearch.com/blog/agentic-ai-commercialization-automotive-industry?utm_campaign=40465778-Kardome%20Voice%20AI%20Authority%20Drive&amp;utm_source=hs_email&amp;utm_medium=email&amp;_hsenc=p2ANqtz-8Tadr-SdEZ_CfIrDKrtKFoiCJ4flK5e-jNbaHhOGHlfMhCGwXB6U4ffDFF4yhoqeOWG2tS">ABI Research projects that vehicle shipments with Agentic AI will grow to nearly 70 million by 2035. The analysis</a> suggests hybrid edge-plus-cloud architectures will be the standard for commercializing agentic AI by then.</li><li>Subscription economics. Software revenue is climbing, but drivers only pay for utility. While ADAS subscriptions are growing (<a href="https://insideevs.com/news/785581/gm-subscription-software-super-cruise/?utm_campaign=40465778-Kardome%20Voice%20AI%20Authority%20Drive&amp;utm_source=hs_email&amp;utm_medium=email&amp;_hsenc=p2ANqtz-8Tadr-SdEZ_CfIrDKrtKFoiCJ4flK5e-jNbaHhOGHlfMhCGwXB6U4ffDFF4yhoqeOWG2tS">GM’s OnStar and Super Cruise set the example</a>), hardware paywalls have faced backlash. Success in voice depends on natural, valuable interactions.</li></ul><h3>Meet Kardome in Taipei!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ORw4okq6eJWXIkWT" /></figure><p>Meet Kardome at InnoVEX in Tapei</p><p>Kardome will be at InnoVEX, June 2–5, 2026, booth S0905, Taipei Nangang Exhibition Center, Hall 2 (TaiNEX 2). If you’ll be there, <a href="https://hubs.ly/Q04gw69v0">book time with</a> <a href="https://hubs.ly/Q04gw69v0">our CEO onsite</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=37960d07ea29" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2026 Voice AI Trends: Engineering the Interface of the Future]]></title>
            <link>https://medium.com/@kardome/2026-voice-ai-trends-engineering-the-interface-of-the-future-8b2834cca600?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/8b2834cca600</guid>
            <category><![CDATA[automotive]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[hybridai]]></category>
            <category><![CDATA[ai-on-device]]></category>
            <category><![CDATA[ces-2026]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Sun, 01 Feb 2026 14:06:24 GMT</pubDate>
            <atom:updated>2026-02-02T22:05:55.624Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J27fsThdU5UZ_fjGII6D6Q.jpeg" /></figure><p>The mantra remains true: Voice is the natural interface for human-machine interaction, the true successor to buttons and touchscreens. However, engineers and product managers at major OEMs understand the uncomfortable reality: Despite decades of investment, current voice AI systems are failing the “real world” test. According to <a href="https://www.gartner.com/en/newsroom/press-releases/2025-10-20-gartner-identifies-the-top-strategic-technology-trends-for-2026">Gartner</a>, nearly half of AI models never make it to production because they cannot handle the messy reality of daily use — noisy cars, chaotic smart homes, and busy factory floors. In these environments, legacy cloud-centric systems perform barely better than 1990s technology.</p><p>In 2026, that gap between expectation and reality will no longer be acceptable. Three structural trends will separate legacy voice UI from production-ready systems:</p><ol><li><strong>Hybrid Voice AI</strong> — on device -first, cloud-augmented architectures instead of cloud-centric pipelines</li><li><strong>Spatial Awareness</strong> — 3D acoustic scene understanding and robust multi-speaker separation</li><li><strong>Cognition AI</strong> — moving from command-based interfaces to context-aware conversational agents</li></ol><p>Taken together, these trends make one thing clear: Achieving seamless, ubiquitous voice interaction is not about adding another feature; it requires re-architecting the software stack. Today’s cloud-centric, LLM-heavy pipelines are fundamentally too slow, too expensive to keep always on, and too detached from local context for reliable daily use.</p><p>High-fidelity perception and rapid decision-making must run on the device processor, with the cloud reserved for long-horizon reasoning and large-context tasks.</p><p><strong>The Flaw in Legacy Architecture and the Latency Imperative</strong></p><p>The primary architectural limitation of current far-field voice recognition systems is their simplistic method of sound scene analysis. Most employ acoustic arrays to determine a sound’s <strong>direction of arrival (DOA)</strong>.</p><p>In acoustically complex environments, which are the default operating conditions, reverberation causes sound waves to bounce off surfaces, creating a “hall of acoustic mirrors”. To a DOA-only system, a single speaker appears as hundreds of different sound sources arriving simultaneously, making it impossible to accurately decode the soundscape and resulting in poor reliability. When multiple speakers are present, these reflections interact and overlap, making separation impossible and the system’s behaviour even less predictable.</p><p>Furthermore, the cloud-only approach introduces fatal latency for conversational systems. Since humans pause approximately 200 milliseconds between conversational turns, systems relying on 1–3 second cloud round-trips for every utterance cannot deliver natural conversation; they deliver walkie-talkie chat. For automotive safety, robotics control, and industrial automation, this latency is unacceptable — decisions must be near-instantaneous.</p><p>These two constraints, poor spatial modeling on device and cloud-bound latency in the back end, will force OEMs toward hybrid voice AI architectures that put robust spatial awareness and fast decision-making on device, with the cloud used selectively rather than by default.</p><h3>1. Hybrid Voice AI: The Architecture of 2026</h3><p>The breakthrough architecture that delivers the required speed, context, and reliability mimics human cognition by using a dual system that splits intelligence between fast, local reflexes and slow, deliberate reasoning. This <strong>Hybrid Voice AI Architecture</strong> will serve as the reference design for OEMs looking to move voice interaction from rigid command-and-control to fluent, context-aware dialogue.</p><p>This architectural pivot is not unique to voice; it mirrors the broader “AI Infrastructure Reckoning” highlighted in<a href="https://www.deloitte.com/us/en/insights/topics/technology-management/tech-trends.html"> Deloitte’s Tech Trends 2026</a>, which reports that industries are aggressively shifting from “cloud-first” to “strategic hybrid” — using the cloud for elasticity and the devices for the immediacy that real-time interaction demands.</p><p><strong>The Reflex Layer: Device AI (System 1)</strong></p><p>This layer consists of high-performance, always-on small language models (SLMs) and processing models embedded directly on dedicated silicon (NPUs or dedicated AI accelerators).</p><p>System 1 handles acoustic perception and the immediate execution of simple commands (e.g., “Lights on”) locally, with near-zero latency, managing roughly <strong>80% of daily interactions</strong> without ever requiring a cloud round-trip.</p><p>This architecture delivers multiple critical advantages: it dramatically enhances <strong>data privacy</strong>, as sensitive voice data never leaves the device; it enables <strong>accurate 3D voice capture</strong> through dedicated acoustic processing pipelines that can perform spatial audio analysis, and multi-speaker localization directly on-chip; and it provides a <strong>reliable, always-responsive interface</strong> with consistent sub-200ms response times, independent of network conditions, server load, or internet availability.</p><p><strong>The Reasoning Layer: Cloud LLMs (System 2)</strong></p><p>This system’s prefrontal cortex layer is activated only when the device determines that complex reasoning, deep knowledge retrieval, or creative generation is necessary. This resource conservation addresses the prohibitive cost of running heavyweight LLMs continuously.</p><p>This split-system approach is rapidly becoming the industry standard.<a href="https://www.gartner.com/en/newsroom/press-releases/2025-10-20-gartner-identifies-the-top-strategic-technology-trends-for-2026"> Gartner’s Top Strategic Trends for 2026</a> predicts that “Hybrid Computing” adoption will surge to 40% by 2028, as enterprises realize that pure cloud models cannot support the economic or performance requirements of complex, always-on AI.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/289/1*yAy8hVYu3EWsJFDVAseoiQ.jpeg" /></figure><h3>2. The Necessity of Spatial Awareness for Context-Aware AI</h3><p>The foundation of the 2026 voice stack is not merely improved beamforming; it is <strong>Spatial Hearing AI</strong>, which gives devices the auditory intelligence they need to function in the real world. This proprietary technology moves beyond simplistic DOA to <strong>pinpoint a sound source’s </strong>precise location in 3D space.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*uGB5-ZTB4wwPXzDZ4m0pkw.jpeg" /></figure><p>It solves the challenge of separating a target speaker from many overlapping voices in a noisy, reverberant space by performing multi-dimensional soundscapes analysis.</p><p>Instead of being confused by echoes, the system utilizes the entire reflection pattern a voice creates within a room, treating it as a unique “acoustic fingerprint” for that specific position. The AI passively infers this fingerprint to effectively map the environment.</p><p><strong>The Standard for Noise Reduction in 2026: Source Separation<br></strong>The result of this Spatial Hearing AI is advanced source separation, enabling the device to isolate individual voices in real time, even amid music, traffic, or simultaneous conversations.</p><p>This means the device hears each user as if they were speaking alone in a quiet room. This capability is critical for safety-focused automotive applications and a multi-user smart home system.</p><p>This focus on environmental understanding aligns with<a href="https://www.gartner.com/en/newsroom/press-releases/2025-10-20-gartner-identifies-the-top-strategic-technology-trends-for-2026"> Gartner’s 2026 “Physical AI” trend</a>, which defines the next generation of AI as systems that leave the screen to “actively sense and navigate” the real world. The urgency of this capability is further reflected in the<a href="https://www.thebusinessresearchcompany.com/report/audio-source-separation-artificial-intelligence-ai-global-market-report"> Audio Source Separation AI Market</a> reports, which project over 38% annual growth through 2030 as industries race to solve the “cocktail party problem” in hardware.</p><h3>3. Cognition AI: From Command-Based to Conversational Agents</h3><p>The shift from command-based voice assistants to truly <strong>conversational, context-aware agents</strong> requires more than just clean audio; it requires intelligence layered on top of the clean stream. In the 2026 voice stack, this is the role of<strong> Cognition AI,</strong> a lightweight SLM that runs on-device.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/330/1*Ipd38xheY0Yk8GmAQzFaXQ.png" /></figure><p>Cognition AI is trained to interpret intent and maintain short-term conversational context. It couples Spatial Hearing AI’s spatial information with semantic understanding to determine whether an utterance is directed at the device or is merely an ambient discussion. This ability to distinguish a direct command (“Lights on”) from conversational flow (“Should we turn the lights on?”) is the key to enabling human-level dialogue and allowing systems to follow multi-step instructions without rigid command structures.</p><p>This transition marks the end of the “chatbot” era and the beginning of what<a href="https://www.businesswire.com/news/home/20251023490057/en/IDC-FutureScape-2026-Predictions-Reveal-the-Rise-of-Agentic-AI-and-a-Turning-Point-in-Enterprise-Transformation"> IDC’s FutureScape 2026</a> calls the “Rise of Agentic AI” — where systems stop being passive tools and become proactive teammates that understand workflows and intent.</p><p>This confluence of Spatial Hearing AI (the auditory sensor) and Cognition AI (the contextual interpreter) operating on the device transforms voice stacks from unreliable gadgets into aware and responsive agents, making them ready for integration into the emerging <strong>Physical AI</strong> ecosystem. The next 24 months will see this transition accelerate, establishing the Hybrid Voice AI Architecture as the expected standard.</p><p><strong>2026 Voice AI Trends OEMs Must Design For</strong></p><p>To lead the voice-first transition in 2026, OEMs must fundamentally pivot their architecture away from cloud dependency and integrate distributed intelligence into their platforms:</p><p><strong>Trend #1 — Hybrid Voice AI (Device-First, Cloud-Augmented)<br></strong> • <strong>Prioritize Device Compute</strong>: The future of voice reliability, privacy, and speed hinges on lightweight, high-performance models (Spatial Hearing AI and Cognition AI) running continuously on-device.<br> • <strong>Software-First Performance on Existing Devices <br></strong>With efficient Spatial Hearing AI and SLM implementations, OEMs can achieve near-zero-latency voice interaction on the device using the CPUs, DSPs, and NPUs they already ship — reducing the need for additional hardware while still meeting strict safety and UX requirements.<br> • <strong>Adopt Hybrid Architecture for Cost and Privacy</strong>: The Hybrid Voice AI Architecture minimizes expensive cloud LLM usage and ensures that sensitive voice data, especially for the 80% of trivial requests, never leaves the device, eliminating the trade-off between performance and superior privacy.</p><p><strong>Trend #2 — Spatial Awareness — Mandate 3D Acoustic Mapping</strong>: Legacy DOA technology is obsolete. New systems must incorporate multi-dimensional soundscape analysis and acoustic fingerprinting to achieve reliable source separation in chaotic acoustic environments.</p><p><strong>Trend #3 — Cognition AI (From Accuracy to Context Intelligence)<br></strong>From Accuracy to Contextual Intelligence: True reliability transcends simple command recognition. The system must maintain rich contextual awareness across multiple dimensions: identifying who is speaking through voice biometrics, determining where they are in 3D space via acoustic localization, inferring intent by distinguishing direct commands from ambient conversation, and preserving conversational memory of recent dialogue history. This contextual understanding transforms the interaction paradigm from rigid, isolated commands to natural, fluid dialogue where follow-up questions, pronouns, and implicit references work seamlessly — much like human conversation.</p><p>Taken together, these trends define what it actually means to treat voice as the primary interface of 2026. OEMs that internalize them — building Hybrid Voice AI, Spatial Hearing AI, and Cognition AI into the core of their platforms — won’t just ship better assistants; they’ll set the benchmark for how humans expect machines to listen and respond in the years ahead.</p><p>This is what it means to engineer voice not as a feature, but as the primary interface of 2026</p><p><strong>Want to learn more about Kardome Voice AI?</strong></p><p>Book your demo here : <a href="https://www.kardome.com/request-a-demo/">https://www.kardome.com/request-a-demo/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b2834cca600" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Context-Aware Voice AI Creates Personalized In-Car UX]]></title>
            <link>https://medium.com/@kardome/how-context-aware-voice-ai-creates-personalized-in-car-ux-a6b68e9f7f16?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/a6b68e9f7f16</guid>
            <category><![CDATA[voice-tech]]></category>
            <category><![CDATA[mobility]]></category>
            <category><![CDATA[context-aware]]></category>
            <category><![CDATA[voice-assistant]]></category>
            <category><![CDATA[voice-ai]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Wed, 22 Oct 2025 15:28:33 GMT</pubDate>
            <atom:updated>2025-10-22T15:28:33.365Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*A7xnyRpDgSnxVStnuNC0eA.png" /><figcaption>Software is driving the user experience, and context-aware voice AI is revolutionizing it by creating custom in-car experiences. This shift moves beyond rigid voice assistants to truly personalized interactions, boosting engagement, safety, and brand loyalty for automotive OEMs.</figcaption></figure><p>In an era where software defines the driving experience, how are you making sure your vehicles not only keep up but also lead the way in personalized driver engagement? And I’m beyond just talking about personalized settings, but rather a personal experience with the voice assistant — so natural, it feels like you’re talking to a close friend, rather than a machine. The stakes for automotive brands are higher than ever: overlooking this gap risks losing customer loyalty and falling behind in a crowded market.</p><p>Think about your digital life today. You can talk to your AI chatbot, which, if you’ve allowed it to save your chat history, knows what’s important to you and responds like a trusted friend or advisor. When you come home and say, ‘I’m home,’ the temperature adjusts to your preference, your favorite song plays, and the lights turn on as you like. It’s seamless, personalized, and frankly, we’ve come to expect it. So, why does the experience in our cars often feel like a step back in time?</p><h3>The Problem: A Disconnected Driving Experience</h3><p>Most in-car voice assistants today are rigid. They treat every request the same, no matter who’s talking, what they’re trying to do, or what’s going on around the car. It <em>feels</em> like you’re talking to a rigid machine. This rigidity often leads to frustrating moments, unnecessary distractions, and missed opportunities for automotive OEMs to truly connect with their customers, ultimately leading to disengagement and abandonment of the voice interface.</p><p>The truth is, making the in-car user experience personal isn’t just a fancy extra anymore. It’s essential for car manufacturers who want to stand out, drive the adoption of in-car features, and foster enduring brand loyalty.</p><h3>Overcoming the Hurdles to Truly Personalized Experiences</h3><p>The initial hurdle of personalization, simply selecting a profile, remains surprisingly cumbersome in many vehicles today. Drivers often contend with touchscreens, diverting their attention from the road to tap through menus. Bluetooth connections, while convenient for a single user, usually fail when multiple registered drivers are in the vehicle together, resulting in incorrect profile activation. And relying on key fobs for automatic profile loading is equally flawed; families frequently share key fobs, or the spare is tucked away for safekeeping, negating any potential for seamless, individualized settings.</p><blockquote><em>For automotive OEMs, investing in context-aware voice AI isn’t just about enhanced user experience; it’s about achieving measurable business outcomes and securing a competitive edge in a rapidly evolving market.</em></blockquote><p>These friction points undermine the very promise of a personalized in-car experience, leaving users frustrated before their journey even truly begins.</p><p>Another barrier to natural, seamless in-cabin voice interaction is the need for wake words and rigid commands, which requires you to remember the exact command phrase. Remove the need for wake words and add natural command inference that understands your intent, and you genuinely make interactions feel as personal as talking to a friend.</p><h3>What Exactly is a “Context-Aware” Voice Assistant?</h3><p>A context-aware voice user interface (VUI) is a voice assistant that can gather and utilize various pieces of information to provide more relevant and personalized responses. It goes beyond just understanding the words you say; it figures out the “who, what, where, and when” of your conversation. To create truly intelligent systems, multiple contextual elements must work together, including:</p><ul><li><strong>Who’s Talking </strong>(Speaker Identification): Knowing who you are is the foundation for personalization. This allows the system to access your individual profile, preferences, and other details, such as your age or preferred language.</li><li><strong>Where You Are in the Car</strong> (Speaker Position): Knowing which seat the speaker is talking from helps form the correct responses and actions to the right person.</li><li><strong>How You’re Doing</strong> (Driver State): Recognizing if the driver is alert or tired can trigger helpful safety suggestions.</li><li><strong>What Else is Happening</strong> (Ambient Audio): Detecting sounds outside the car, like emergency sirens, can significantly boost safety.</li></ul><h3>Examples of how context awareness changes in-car UX:</h3><ul><li><strong>Personalized Navigation</strong>: When a registered driver says, “Navigate to work,” the car takes them to their saved work address without needing to spell it out. If two people are in a car, and a different registered person says, “navigate to work,” it takes them to their saved work address. The system knows who is giving the command. No confusion, no need to touch a screen.</li><li><strong>Information Just for You</strong>: Imagine driving past a famous building. A passenger asks, “What’s that building?” The system tailors the explanation based on the user’s profile (e.g., historical facts for a history lover, engaging and straightforward for a child, and architectural details for an engineer).</li><li><strong>Proactive Safety:</strong> If the system detects signs that the driver may be getting sleepy (perhaps from their voice patterns or, using a visual mode system, how they’re looking at the road), it can suggest taking a coffee break or point out nearby rest stops.</li><li><strong>Your Music, Instantly</strong>: Say “Play my playlist,” and your preferred music starts right away, based on your identified profile and past listening habits — even if the vehicle profile is set to someone else.</li></ul><p>This type of intelligent interaction builds a great deal of trust and engagement with the car. When the system feels like a genuine companion rather than just a tool, you’re much more likely to use it, making your journey safer, more intuitive, and ultimately, more enjoyable.</p><h3>The ROI of Personalization: Driving Business Outcomes</h3><p>For automotive OEMs, investing in context-aware voice AI isn’t just about enhanced user experience; it’s about achieving measurable business outcomes and securing a competitive edge in a rapidly evolving market.</p><p>Consider these compelling statistics and technical considerations:</p><ul><li>Market Growth &amp; Adoption:<a href="https://betanews.com/2025/08/19/ai-in-car-assistant-market-set-to-more-than-double-by-2031/"> </a>The global AI in-car assistant market is expected to reach <a href="https://betanews.com/2025/08/19/ai-in-car-assistant-market-set-to-more-than-double-by-2031/">USD 15.8 billion by 2031, growing at a 12.8% CAGR</a>. The in-vehicle voice assistant market is projected to reach <a href="https://www.marketgrowthreports.com/market-reports/in-car-voice-assistant-market-114562">$12.6 billion by 2033, with an 18.2% CAGR</a>, and<a href="https://betanews.com/2025/08/19/ai-in-car-assistant-market-set-to-more-than-double-by-2031/"> over 69% of new cars </a>in North America and Western Europe now feature voice control.</li><li>Improved Safety and Compliance: 88% of American drivers regularly use in-car voice features, with over <a href="https://www.statista.com/statistics/957808/us-consumers-using-automotive-voice-assistants-by-frequency/">60% citing enhanced safety</a> as a key benefit</li><li>Customer Satisfaction and Differentiation: Automotive OEMs report a <a href="https://www.statista.com/statistics/957808/us-consumers-using-automotive-voice-assistants-by-frequency/">40% year-on-year increase in customer satisfaction</a> ratings tied to advanced voice recognition systems.</li></ul><h3>Strategic Path to Personalized In-Car UI for OEMs</h3><p>At the end of the day, context isn’t optional; it is the indispensable foundation that turns in-car voice UI into a trusted copilot, making it safer, faster, and more natural. It also needs to incorporate a great deal of context. Audio is a fantastic “collector” of this context — understanding where sounds are coming from and their direction is a vital part of the puzzle.</p><p>For OEMs looking to build truly adaptive, personalized, and safe in-car user experiences, the first step is to clearly define their vision. That clear vision will then dictate the necessary hardware and software. From there, the next critical step is to find the right partners who can integrate all these complex pieces seamlessly. Look for partners with proven expertise in multi-modal AI, deep experience with automotive-grade systems, a strong ability to integrate with your existing platforms, and robust privacy frameworks. The right partner understands the unique demands of the automotive environment and can help you navigate the complexities of integrating advanced AI.</p><p>The future of the automotive experience won’t just be about fancy holographic screens, but about how intimately your vehicle understands and anticipates the needs of its occupants. It’s a journey to deeper brand loyalty and truly intelligent mobility — and it starts with a clear vision and the right partners today.</p><h3>Learn how Kardome Automotive delivers human-like hearing and understanding inside every vehicle with Spatial Hearing AI and Cognition AI</h3><p><a href="https://hubs.ly/Q03Pzp5d0">Watch the Video</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a6b68e9f7f16" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Voice AI Becomes Heart of Automotive and Smart Home Experiences]]></title>
            <link>https://medium.com/@kardome/voice-ai-becomes-heart-of-automotive-and-smart-home-experiences-49f3d4331ed1?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/49f3d4331ed1</guid>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[generative-ai]]></category>
            <category><![CDATA[voice-tech]]></category>
            <category><![CDATA[voice-technology]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Sat, 11 Oct 2025 13:53:14 GMT</pubDate>
            <atom:updated>2025-10-11T13:53:14.432Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kBrPEx8zjqB8fXkA1Lylhg.png" /></figure><p>Video Shows In-Cabin Voice AI Excels in Real-World Conditions</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FDJ-7rA9phnE%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DDJ-7rA9phnE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FDJ-7rA9phnE%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/9ffd1611e086e5b5cfc821c3d68b2d5d/href">https://medium.com/media/9ffd1611e086e5b5cfc821c3d68b2d5d/href</a></iframe><p>Kardome’s recently published video demonstrates how its advanced voice AI solution, Kardome Automotive, delivers a superior voice user interface (Voice UI) in the most demanding, real-world car environments.</p><p>The demonstration, conducted at freeway speeds with windows down and music playing, highlights how Kardome’s Spatial Hearing and Cognition AI technologies enable accurate, human-like interaction that adapts to real driving conditions and passenger dynamics. Crucially, the solution runs entirely on-device, delivering fast, personalized, and private responses without relying on cloud connectivity.</p><p>This robust performance addresses significant concerns for automotive OEMs seeking reliable in-vehicle voice UI solutions. It serves as a strong proof of concept for real-world usability, moving beyond controlled lab conditions.</p><p><em>Learn more about how Kardome Automotive achieves 98% accurate speech recognition with windows down, driving at freeway speeds </em><a href="https://hubs.ly/Q03KJG6N0"><strong><em>here</em></strong></a><strong><em>.</em></strong></p><h3>“Beyond the Hype”: What’s Real (and What Still Needs More Work) in In-Cabin Voice AI</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*70sPeOw_caGkSwtnlWcfTw.png" /><figcaption>Beyond the Hype explores what’s succeeding and what still needs work in automotive voice AI.</figcaption></figure><p><a href="https://www.linkedin.com/company/kardome/"><strong>Kardome</strong></a> CEO <a href="https://www.linkedin.com/in/dani-cherkassky/"><strong>Dani Cherkassky</strong></a>’s recent blog offers a candid look at what’s working and the challenges that persist.</p><p>Highlights:</p><ul><li>The “Companion on Wheels” Concept: Voice, vision, and spatial understanding are becoming central to this concept, enabling systems to know who is speaking, from where, and with what intent.</li><li>Generative AI’s Impact: Generative AI facilitates more natural dialogue and predictive/anticipatory behaviors (e.g., suggesting playlists, route changes) beyond simple reactive commands.</li><li>Persistent Technical Gaps: Significant challenges remain, including reliance on wake words and cloud-based large language models, difficulties in achieving truly proactive and human-like conversational dialogue, and the unforgiving nature of in-cabin acoustics.</li></ul><p>Explore “Beyond the Hype” to see what areas are ripe for investment and what challenges OEMs must solve. <a href="https://www.kardome.com/blog-posts/voice-ai-transforming-in-cabin-ux"><strong>Read it here</strong></a></p><h3>𝗧𝗵𝗲 F𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 V𝗼𝗶𝗰𝗲 𝗔𝗜 𝗶𝘀 B𝘂𝗶𝗹𝘁 𝗼𝗻 C𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻</h3><p>Elevating in-vehicle voice AI innovations truly thrives on collaboration and shared vision. Team Kardome recently joined our partners in Korea to celebrate our successes and explore exciting possibilities ahead. Together, we’re excited to unlock new horizons and push the boundaries of what #voiceAI can achieve.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6Oa-lxrvnzWhFRlet_HiYQ.png" /></figure><h3>Automotive Voice AI News</h3><ul><li>Von der Leyen Calls for “AI First” Strategy in Europe’s Mobility Sector: At Italian Tech Week, EU Commission President <a href="https://www.linkedin.com/in/ursula-von-der-leyen/"><strong>Ursula von der</strong> Leyen</a>proposed a continent-wide push for autonomous systems and AI in mobility. <a href="https://www.reuters.com/business/retail-consumer/eus-von-der-leyen-urges-european-push-ai-driven-cars-2025-10-03/?utm_source=chatgpt.com"><strong>Read more</strong></a></li><li><a href="https://www.linkedin.com/company/ford-motor-company/"><strong>Ford Motor Company</strong></a> aims to include an AI assistant in every vehicle: Ford CEO Jim Farley expressed ambition for a built-in AI assistant across all Ford vehicles, starting with its Blue Oval, in response to competitive pressure and the rising expectation of voice/agent features in cars. <a href="https://fordauthority.com/2025/10/ford-ceo-jim-farley-wants-an-ai-assistant-in-every-blue-oval-car/?utm_source=chatgpt.com"><strong>Read more</strong></a></li><li><a href="https://www.linkedin.com/company/soundhoundai/"><strong>SoundHound AI</strong></a> Rolls Out Generative Voice in Jeep Vehicles: SoundHound’s Chat AI Automotive is now live in select Jeep models in Europe, enabling conversational voice beyond basic commands. <a href="https://investors.soundhound.com/news-releases/news-release-details/soundhound-ai-brings-generative-ai-powered-voice-assistant-jeep"><strong>Read more</strong></a></li></ul><h3>Smart Home Voice AI News</h3><ul><li><a href="https://www.linkedin.com/company/openai/"><strong>OpenAI</strong></a>, Jony Ive Face Tech Hurdles for Screenless AI Device: As OpenAI and Jony Ive face challenges in developing their always-on AI device due to privacy and computing infrastructure issues, the industry is realizing that natural conversation depends on advanced, on-device cognition. This is the very foundation of Kardome’s Spatial Hearing and Cognition AI technologies. <a href="https://www.ft.com/content/58b078be-e0ab-492f-9dbf-c2fe67298dd3"><strong>Financial Times</strong></a><strong> | </strong><a href="https://www.bloomberg.com/news/articles/2025-05-21/openai-to-buy-apple-veteran-jony-ive-s-ai-device-startup-in-6-5-billion-deal?srnd=undefined"><strong>Bloomberg</strong></a></li><li>Google Launches Gemini for Home: <a href="https://www.linkedin.com/company/google/">Google</a>introduced Gemini for Home, which replaces Google Assistant on Nest devices and enhances smart home devices with more conversational, context-aware AI. <a href="https://www.theverge.com/tech/788085/hands-on-google-home-speaker"><strong>Read more</strong></a></li><li>Amazon Unveils Alexa+–Ready Echo Devices: <a href="https://www.linkedin.com/company/amazon/"><strong>Amazon</strong></a>’s new Echo models are built for Alexa+, with custom silicon and sensor fusion platforms for faster, more proactive AI voice interactions. <a href="https://www.reuters.com/technology/amazon-refreshes-device-lineup-alexa-ai-home-security-2025-09-30/?utm_source=chatgpt.com"><strong>Read more</strong></a></li></ul><h3>Executive Insights: Key Trends in Voice AI</h3><ul><li>Production deployments accelerate: SoundHound’s Jeep integration and Ford’s public AI commitment signal that generative, conversational voice is becoming a real differentiator in vehicles.</li><li>Edge, Context, Memory: Pivotal to AI Assistant Success: The challenges faced by OpenAI and Jony Ive’s AI device underscore the need for advanced on-device processing, contextual understanding, and memory in AI assistants.</li><li>AI Monetization and Subscription Models Emerge: Google’s Gemini Live and Amazon’s push of Alexa+ hint at future smart voice ecosystem subscription tiers.</li></ul><h3>Stay at the forefront of Voice + AI in automotive and the smart home:</h3><p>Subscribe to our <a href="https://www.linkedin.com/newsletters/the-voice-tech-newsletter-6948330259272400897/"><strong>newsletter</strong> </a>for monthly insights, product launches, and trend analysis.<br>Ready to elevate your product’s voice UI capabilities? <a href="http://kardome.com/contact-us"><strong>Contact us</strong></a> for a personalized demo or technical overview.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=49f3d4331ed1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How SK Networks and Kardome Are Redefining Smart Home Wellness with Voice AI]]></title>
            <link>https://medium.com/@kardome/how-sk-networks-and-kardome-are-redefining-smart-home-wellness-with-voice-ai-379d8cdf2d8c?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/379d8cdf2d8c</guid>
            <category><![CDATA[voice-technology]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[consumer-electronics]]></category>
            <category><![CDATA[ai-robot]]></category>
            <category><![CDATA[automated-smart-homes]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Wed, 17 Sep 2025 13:37:20 GMT</pubDate>
            <atom:updated>2025-09-17T13:37:20.615Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cUAz3ycJINWb2wGvo4f-Tw.png" /></figure><p>Powered by Kardome’s Spatial Hearing AI, SK Networks’ NAMUHX air-purifier robot launches this month. Designed as a true ‘wellness robot,’ NAMUHX combines clean air with intelligent, natural voice interaction.</p><p>When SK Networks, along with appliance affiliate SK Magic, set out to build <em>NAMUHX</em>, its innovative air purifier robot, it wasn’t simply about creating a smart appliance. It was about reimagining how technology could interact with people in the home: listening, understanding, and adapting to human needs in real time. To bring that vision to life, SK partnered with Kardome, a pioneer in Spatial Hearing AI.</p><p>Launching in Korea this September, with plans to expand into the U.S. and Malaysia by December, NAMUHX is designed as a “wellness robot,” combining air purification with intelligent voice interaction. NAMUH is HUMAN spelled backwards, emphasizing the company’s human-centric approach to its technology.</p><p>While looking to integrate a large language model solution for its robot purifier, SK Manager Yongjae Choi said, “We found Kardome’s Spatial Hearing AI and Cognition AI solution capabilities to be incredibly appealing.”</p><p>Key components of Kardome’s Spatial Hearing used by SK are the audio front-end and direction of arrival models, while from the Cognition AI package, SK is using Kardome’s Wake Detection model.</p><p>‍</p><blockquote><em>“The ability to handle constant signal changes from a moving robot with Kardome’s Spatial Hearing technology is a key advantage.”</em></blockquote><p>‍</p><p>The company’s choice was crucial to its vision of an AI-focused wellness robot that can truly speak, listen, and understand the human voice.</p><p>What makes NAMUHX stand apart from traditional smart devices is its ability to interact with users in dynamic environments effectively. Unlike classical signal processing methods, such as beamforming, which struggle when a device is moving, Kardome’s Spatial Hearing solution, powered by AI, enables the robot to process constantly changing audio input smoothly, ensuring its performance shines even with NAMUHX on the move.</p><p>“The ability to handle constant signal changes from a moving robot with Kardome’s Spatial Hearing technology is a key advantage,” Choi said.</p><p>NAMUHX also uses Kardome’s Wake Detection, enabling the robot to respond to its own custom wake word, <em>“Hi NAMUH.”</em> This feature allows brands, and ultimately end users, to define unique wake words for their devices, enabling more personalized interaction and brand distinction.</p><h3>Spatial Hearing AI: The Technology Behind the Experience</h3><p>At the core of this partnership is Kardome’s industry-leading Spatial Hearing AI solution, which continuously maps the acoustic environment, separates voices from background noise, and assigns each part of a conversation to the correct speaker.</p><p>It allows devices not just to detect the direction of sound but also to pinpoint its exact location, improving accuracy, speed, and responsiveness while keeping all processing on the device for privacy.</p><p>This advanced feature results in a superior user experience for NAMUHX customers, allowing seamless voice control even as its devices move around the home, ultimately boosting product satisfaction and reducing support issues caused by voice command errors.</p><h3>Why SK Chose Kardome</h3><p>SK’s rigorous evaluation process revealed a critical gap in conventional voice AI solutions for moving robots. As Manager Choi explains, “With conventional beamforming alone, it would have been very difficult to develop a robot that can understand users while moving 360 degrees.”</p><p>This fundamental challenge, common across the smart device industry, is precisely what Kardome’s Spatial Hearing AI was engineered to overcome, making it the clear choice for NAMUHX.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FdhNxgCCUmd8%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DdhNxgCCUmd8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FdhNxgCCUmd8%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/c9874eef23a19a4ca104e43e6a55a2bc/href">https://medium.com/media/c9874eef23a19a4ca104e43e6a55a2bc/href</a></iframe><h3>The Future of Wellness: Voice Biometrics and Cognition AI</h3><p>SK’s vision for wellness robots extends far beyond current capabilities. They are committed to integrating advanced voice AI into their line of smart wellness devices for comprehensive health management and user behavior analysis.</p><p>Today, NAMUHX combines air purification with vision AI to analyze health conditions; for example, measuring blood pressure or blood oxygen through facial recognition. But SK has tasked Kardome with exploring an even more ambitious frontier: using voice as a health signal.</p><p>By analyzing subtle cues in speech — tone, cadence, or vocal biomarkers — future wellness robots could detect stress, fatigue, or emotional state, providing a richer picture of user health. This ties directly into SK’s focus on personalized wellness management.</p><p>Looking ahead, SK is particularly enthusiastic about Kardome’s Voice ID<strong> </strong>product, which would allow robots to distinguish between individuals and deliver tailored experiences.</p><p>The company is also exploring Kardome’s Cognition AI solution, which would go beyond recognition to interpret user intent and emotion. Together, these technologies could transform NAMUHX and future devices into genuine companions, robots that don’t just respond to commands but understand context, mood, and personal needs.</p><p>“We are interested in analyzing voice data more deeply to understand user intent and emotions, exploring new features and solutions with Kardome’s voice AI,” Choi said.</p><h3>A Partnership Built on Collaboration</h3><p>While technology is at the heart of the partnership, both teams highlight the collaboration itself as a differentiator.</p><p>Said Choi, “Kardome’s R&amp;D team has shown a level of dedication we rarely see. They brainstorm solutions with us daily and have been deeply committed to making this product a success.”</p><p>Kardome CEO <strong>Dani Cherkassky</strong> echoed the sentiment: “Working with SK has been incredibly rewarding. Their vision for wellness robotics aligns perfectly with our mission to make voice AI truly intelligent and human-centric. This partnership demonstrates how combining innovation with close collaboration can bring groundbreaking products to life.”</p><p>For OEMs in the consumer electronics industry, this partnership offers a glimpse of what’s possible when advanced voice AI meets forward-thinking product design. It’s not simply about voice commands — it’s about creating devices that understand, adapt, and connect with people in ways that feel natural and intuitive.</p><p>As SK and Kardome continue to push the boundaries of what voice AI can do, NAMUHX stands as a powerful proof point: the future of consumer technology will be conversational, contextual, and deeply personal.</p><h4><a href="https://www.kardome.com/landing/kardome-spatial-hearing-ces">Click here</a> to learn more about Kardome’s Spatial Hearing AI technology</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=379d8cdf2d8c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Beyond the Hype: How Voice AI is Transforming In-Cabin UX and Where It Still Falls Short]]></title>
            <link>https://medium.com/@kardome/beyond-the-hype-how-voice-ai-is-transforming-in-cabin-ux-and-where-it-still-falls-short-08ce0d73886e?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/08ce0d73886e</guid>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[voice-recognition]]></category>
            <category><![CDATA[speech-recognition]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[spatial-hearing]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Thu, 21 Aug 2025 16:31:32 GMT</pubDate>
            <atom:updated>2025-08-21T16:31:32.962Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CA_GqAlwPDNMkb-nSs-wmg.png" /><figcaption><em>Voice AI promises a seamless, natural in-car experience, yet its full potential remains untapped despite significant breakthroughs. The question is no longer if voice will be central to the in-cabin experience — it’s whether the industry is prioritizing the right problems to make that experience truly intelligent.</em></figcaption></figure><p>Voice AI promises a seamless, natural in-car experience, yet its full potential remains untapped despite significant breakthroughs. The question is no longer <em>if</em> voice will be central to the in-cabin experience — it’s whether the industry is prioritizing the right problems to make that experience truly intelligent.</p><p>This article explores the current capabilities and limitations of in-car voice AI, identifies critical areas for advancement toward the “companion on wheels” vision, and argues for immediate, focused innovation that meets escalating consumer expectations, while capturing a rapidly expanding market opportunity.</p><h3>An HMI Roadmap for OEMs: The Shift to Proactive Intelligence</h3><p>The automotive industry is moving beyond buttons and touchscreens towards personalized, intelligent, multimodal experiences — with voice and vision as the primary modalities for natural Human Machine Interaction (HMI).</p><p>A genuine “companion on wheels” demands more than basic voice commands. It requires comprehending the entire in-cabin acoustic scene — not just <em>what</em> was said, but <em>who</em> said it and <em>from where</em> — alongside deep contextual understanding for proactive assistance. Achieving this level may also require integrating audio with visual cues, such as eye-tracking.</p><p><a href="https://www.smarteye.se/smart-eye-at-ces-2025/">Smart Eye’s CES demonstration of a 3D interior sensing system</a>, using eye-tracking and cabin monitoring, exemplifies the industry’s move toward proactive, intelligent in-vehicle companions.</p><p>‍</p><blockquote><em>With the in-car voice technology market projected to exceed </em><a href="https://www.precedenceresearch.com/automotive-voice-recognition-system-market?utm_source=chatgpt.com"><em>$13B by 2034</em></a><em>, OEMs that lag on user experience will lose market share not to “better cars,” but to better </em>conversations<em>.</em></blockquote><p>‍</p><h3>Cutting Through the Hype: What Voice AI Can Deliver Now</h3><p>In-car voice AI has moved light-years beyond simple commands like “Call Mom” or “Turn on the radio.” Both consumer demand for seamless, hands-free interaction and rapid innovation drive the shift. Key trends accelerating this transformation include:</p><ol><li><strong>The Rise of Generative AI</strong></li></ol><ul><li><strong>More Natural Conversation and Intent Understanding:</strong> Generative AI is a game-changer. Modern voice assistants are now engaging in more complex and natural conversations. They can understand context, follow up on previous requests, and handle multi-layered commands like, “Find a coffee shop near me with free Wi-Fi and outdoor seating.”<strong>‍</strong></li><li><strong>Predictive and Proactive Intelligence:</strong> AI-driven VUIs are becoming predictive, learning drivers’ habits to offer proactive assistance, such as the car voice assistant automatically suggesting a preferred route or playlist based on the time of day, or pre-emptively warning about vehicle issues.</li></ul><ol><li><strong>Brand Personalization as a Differentiator:</strong></li></ol><ul><li>Automakers are increasingly developing their own branded voice assistants to create a unique and consistent brand experience. <a href="https://www.mbusa.com/en/future-vehicles/mbux-hyperscreen">Mercedes-Benz with MBUX</a>, <a href="https://www.nio.com/blog/nomi-worlds-first-vehicle-artificial-intelligence">Nio with its NOMI chatbot</a>, and others are building systems that go beyond generic voice commands to provide a more integrated and personalized interaction, strengthening brand loyalty</li></ul><ol><li><strong>Overcoming Core Technical Challenges for Scalability</strong></li></ol><ul><li><strong>Advanced Noise Cancellation:</strong> Advanced noise cancellation and multi-microphone arrays to create “acoustic zones” within the noisy cabin, ensuring accurate command understanding by filtering background noise.</li><li><strong>Hybrid On-Device and Cloud Processing for Responsiveness:</strong> Hybrid on-device and cloud processing addresses connectivity and response time. Essential commands are processed on the edge for speed, while complex requests go to the cloud, ensuring responsiveness even without Internet.</li></ul><ol><li><strong>Elevating Safety and Accelerating User Adoption</strong></li></ol><ul><li><strong>Minimizing Distraction and Enhancing Safety:</strong> A primary goal is enhanced safety by reducing “eyes off the road” time. Replacing physical buttons and complex touchscreen menus with voice commands lowers cognitive load, keeping hands on the wheel.</li><li><strong>Bridging the Trust Gap Through Performance:</strong> Earning user trust remains a challenge due to accuracy issues, privacy concerns (always-on microphones), and frustrating experiences with older, less capable systems. Automakers are addressing this with clearer privacy policies and demonstrably reliable systems.</li></ul><h3>Where Voice AI Still Falls Short and How We Can Fix It</h3><p>Voice UIs have made significant progress in mimicking human conversation, often sounding incredibly natural. However, their interaction falls short of human conversation, struggling with ambiguity, sarcasm, humor, and unstructured dialogue. There are critical architectural and acoustic limitations that prevent them from truly replicating natural human conversation and providing seamless, proactive assistance. These include:</p><h4>Architectural Limitations: The Wake Word Problem</h4><p>A fundamental barrier to natural voice interaction is the wake word, which forces a rigid, turn-based interaction rather than spontaneous human-like conversation, frequently leading to frustrating and out-of-context exchanges.</p><p>This constraint stems from the current voice UI architecture: a small, on-device software detects the wake word, opening a “gate” that connects to a cloud-based large language model (LLM) for processing. Hypothetically, this “gate” to the cloud LLM could be left open, making wake words unnecessary. However, continuous cloud LLM processing is impractical due to prohibitive costs and significant privacy concerns (i.e., constant audio uploading to the cloud).</p><p>Overcoming these limitations demands a fundamental architectural paradigm shift: the development and deployment of smaller, purpose-built Small Language Models (SLMs) that run directly on the device. Enabled by advances in edge computing, an always-on, always-listening SLM could enable seamless, human-like voice interaction without constant cloud reliance, and would eliminate the need for wake words.</p><h4>Conversational Flow and Proactive Intelligence</h4><p>Human conversation is a complex exchange of turn-taking, interruptions, and non-verbal cues. Current voice UIs struggle with this. A slight delay in response disrupts the natural rhythm of a conversation. If you interrupt a person, they can adapt. If you interrupt a Voice UI, it might get confused, reset, or fail to process the new command entirely. A human might also spontaneously show empathy by saying, “You sound upset, is everything okay?” based on your tone of voice — an AI won’t.</p><p>The always-on SLM architecture previously described would be pivotal in enabling the VUI to understand not just conversational context but also critical subtext and emotional cues, paving the way for truly predictive, proactive, and genuinely human-like interactions.</p><h4>The Unforgiving Nature of in-Cabin Acoustics</h4><p>Current in-car voice UI faces specific and significant challenges due to the unique acoustic environment of a vehicle. These challenges include:</p><ul><li><strong>Reverberation and Echo:</strong> The enclosed space of a car cabin causes sounds, including the driver’s voice, to bounce off hard surfaces like windows and dashboards. This creates echoes and reverberation, which can corrupt the original speech signal and make it difficult for the voice AI to transcribe the original speech accurately.</li><li><strong>Competing Speakers and Crosstalk:</strong> With multiple people in the car, the system must differentiate between the person giving a command and other passengers talking. This “crosstalk” can confuse the system, leading to incorrect responses or a complete failure to understand the command. Traditional systems often struggle to isolate a single speaker from a mix of conversations.</li><li><strong>Ambient Noise:</strong> Cars are inherently noisy environments. Road noise, engine hum, air conditioning fans, music playing, and open windows all contribute to a high level of background noise that can overwhelm the user’s voice and degrade recognition accuracy.</li></ul><p>Advanced audio processing techniques, such as Spatial Hearing AI, provide solutions:</p><ul><li><strong>Isolating individual speakers</strong> by clustering speech signals based on location</li><li><strong>Reducing reverberation and noise</strong> by locating speakers accurately</li><li><strong>Integrating seamlessly</strong> with existing microphone arrays as software, avoiding complex, multi-microphone setups.</li></ul><p>These methods produce clear, isolated speech signals, enhancing accuracy and reliability.</p><h3>Why These Challenges Matter Now</h3><ul><li>Poor performance <strong>hurts </strong><a href="https://www.jdpower.com/"><strong>JD Power</strong></a><strong> and UX scores</strong>, which in turn affect brand loyalty and repurchase intent.</li><li>With the in-car voice technology market projected to exceed <a href="https://www.precedenceresearch.com/automotive-voice-recognition-system-market?utm_source=chatgpt.com">$13B by 2034</a>, OEMs that lag on user experience will lose market share not to “better cars,” but to better <em>conversations</em>.</li><li>Regulatory pressures around biometrics, data handling, and <a href="https://www.eetimes.eu/voice-technology-answering-europes-call-for-safer-automotive-controls">driver distraction</a> are only increasing. The next five years will be decisive.</li></ul><h3>Conclusion: Beyond the Hype — Realism and Opportunity</h3><p>While in-car Voice AI has advanced significantly with Generative AI and improved noise cancellation, its full potential remains untapped. Despite progress in natural conversation and personalization, challenges persist, including reliance on wake words, difficulty with human conversational nuances, and demanding car acoustics. The future of in-car voice AI depends on smaller, on-device SLMs and advanced spatial audio, enabling proactive, predictive systems for a “companion on wheels” experience. For OEMs, this means market differentiation, enhanced user experience, and privacy compliance. Success will be measured by tangible improvements in safety, convenience, and user satisfaction, not hype.</p><h4>About Dani Cherkassky</h4><p>Dani Cherkassky, PhD, is the CEO of Kardome, a voice AI company offering <a href="https://www.kardome.com/spatial-hearing">Spatial Hearing AI</a> and Cognition AI technologies that unlock hands-free, frustration-free voice-led experiences in any real-world condition, from noisy car cabins to shared living spaces, and more.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=08ce0d73886e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Voice Tech’s Rush Hour: Navigating the Chaos of In-Car UI and the Future of Spatial Hearing]]></title>
            <link>https://medium.com/@kardome/voice-techs-rush-hour-navigating-the-chaos-of-in-car-ui-and-the-future-of-spatial-hearing-5bf4751d041e?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/5bf4751d041e</guid>
            <category><![CDATA[spatial-hearing]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[mobility]]></category>
            <category><![CDATA[voice-tech]]></category>
            <category><![CDATA[autotech]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Mon, 28 Jul 2025 01:49:12 GMT</pubDate>
            <atom:updated>2025-07-28T01:49:12.612Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rQCqVBcEJ9jNUVzM0yIEaw.jpeg" /><figcaption>Is your voice assistant ready for rush hour? This latest edition of the Voice Tech newsletter dives into the rugged reality of voice tech in the automotive industry, showcasing Kardome’s breakthrough demonstrations at AutoTech and the key developments shaping the voice user interface (voice UI) landscape today.</figcaption></figure><h3>Takeaways from AutoTech</h3><p>The ideal vision of voice technology often involves quiet spaces and flawlessly responding personal assistants. However, the actual test of voice UI lies in the challenging real-world environment, especially in vehicles characterized by engine noise, multiple voices, and unpredictable road disturbances.</p><p>Kardome’s recent AutoTech demo represented a significant milestone. We provided automotive OEMs with an immersive demonstration of our <a href="https://www.kardome.com/spatial-hearing">Spatial Hearing AI technology</a> under simulated, extremely high-level road noise conditions.</p><p>Even within an intentionally disruptive acoustic setting, <a href="https://www.linkedin.com/company/kardome/">Kardome</a>’s voice UI software consistently isolated, identified, and comprehended individual speakers regardless of location, providing clear signals to the vehicle’s speech recognition system.</p><p>The demo impressed engineers and UX leaders, highlighting that reliable spatial hearing is no longer a supplementary feature but necessary for intent-based, in-car voice UI that functions effectively for the driver and all passengers in real-world scenarios.</p><p>We also presented our partnership with <a href="https://www.linkedin.com/company/soundhoundai/">SoundHound AI</a> through the integration of their Automotive Chat AI feature. This combined system provides natural audio like a human, accurately identifying speakers, eliminating distractions, understanding speech, and delivering precise, timely responses to both the driver and passengers, even in the third row. &gt;&gt;&gt; <a href="https://www.kardome.com/blog-posts/kardome-soundhound-ai-team-up-transform-in-car-voice-ui-multi-zone-conversational-ai">Learn more about the Kardome and SoundHound AI collaboration.</a></p><h3>🗝️Key Partnerships and Integrations</h3><ul><li>LG + Kardome: <a href="https://www.linkedin.com/company/lg-electronics/">LG Electronics</a> has chosen Kardome’s Spatial Hearing AI to drive its next-gen smart TVs, with plans to expand into smart appliances, audio products, and vehicles. By integrating this technology, consumer electronics can now mimic human hearing, delivering smooth, accurate voice UIs even in noisy, multi-speaker environments. &gt;&gt;&gt; <a href="https://www.forbes.com/sites/johnarcher/2025/05/13/lg-oled-and-lcd-tvs-are-about-to-get-easier-to-use/">Read more on Forbes.</a></li><li>Kardome + <a href="https://www.linkedin.com/company/nvidia/">NVIDIA</a>: Kardome’s tech is now integrated with NVIDIA’s DRIVE AGX platform, delivering precise voice isolation, speaker identification, and improved speech recognition, enhancing the in-cabin voice experience for next-gen vehicles. &gt;&gt;&gt; <a href="https://www.kardome.com/blog-posts/kardome-mobility-nvidia-agx-spatial-hearing-ai-voice-recognition">Learn more</a></li></ul><h3>🚘Voice UI and European Safety Standards</h3><p>As vehicles pack more features, the risk of driver distraction increases. In a recent <em>EE Times article</em>, Kardome CEO Dani Cherkassky discusses how voice technology is evolving from a convenience feature to a crucial safety solution. With European regulators focusing more on reducing cognitive load and minimizing eyes-off-road time, in-cabin voice interfaces are becoming a key factor in safer driving, helping drivers stay focused, react more quickly, and meet growing safety requirements standards. &gt;&gt;&gt; <a href="https://47487960.fs1.hubspotusercontent-na1.net/hubfs/47487960/News/EE%20Times%20Europe%20Dani%20on%20Voice%20UI.pdf">Read the article here.</a></p><h3>🌐 Industry Trends and Highlights:</h3><ul><li>Apple at WWDC 2025: Apple unveiled iOS 26 at WWDC, introducing the sleek “Liquid Glass” design and powerful on-device AI features like Live Translation, Hold Assist, and enhanced Shortcuts automation. While impressive, Apple delayed its Siri overhaul until 2026 due to quality concerns. 🔎 Why it matters: <a href="https://www.linkedin.com/company/apple/">Apple</a>’s bet on private, on-device voice AI is a wake-up call to OEMs — users expect voice systems to be fast, local, and secure. However, even Apple’s challenges show how difficult it is to deliver consistent, context-aware voice UX. Source → <a href="https://www.theverge.com/news/685106/apple-ai-siri-delay-craig-federighi-greg-joswiak-wwdc-2025">The Verge</a></li><li><a href="https://www.linkedin.com/company/google/">Google</a> Gemini 2.5: Gemini 2.5 introduced live, interactive voice with emotion recognition, multi-voice TTS, and real-time audio APIs. These tools offer developers granular control over voice personality and interaction. 🔎 Why it matters: This is a blueprint for OEMs looking to create branded voice assistants that feel distinctly “theirs” — and more human. Source → <a href="http://blog.google/">blog.google</a></li><li>Voice AI Startup Surge: In 2024, funding for voice AI startups surged 8× to $2.1 billion, according to <a href="https://www.linkedin.com/company/pymnts-com/">PYMNTS</a>. Startups like <a href="https://www.linkedin.com/company/elevenlabsio/">ElevenLabs</a> and <a href="https://www.linkedin.com/company/openai/">OpenAI</a> are delivering lifelike AI voice agents capable of handling customer support, sales, and service at scale. 🔎 Why it matters: The democratization of high-quality voice tech opens doors for OEMs of all sizes to build custom, reliable voice systems into their products. Source → <a href="https://www.pymnts.com/artificial-intelligence-2/2025/voice-ai-funding-surges-8x-as-businesses-humanize-chatbots/?utm_source=chatgpt.com">PYMNTS</a></li></ul><h3>🔮Future Outlook for Voice AI</h3><ul><li>Real-World Ready: Spatial hearing and understanding are no longer optional — it’s foundational for voice tech that works where we live, drive, and move.</li><li>Privacy-First: On-device AI is critical for fast, secure, and trusted in-vehicle interactions.</li><li>Personalization Wins: Customizable voices and UX are becoming key to brand differentiation.</li><li>Startup-Driven Innovation: With funding booming, expect more rapid, affordable voice integration across industries.</li></ul><h3>❓What’s Your Take?</h3><p>From overcoming noisy environments to driving safer experiences, voice AI is rapidly evolving. What trend will be most influential in 2025 and beyond? Let us know your thoughts in the comments! To keep up with these advancements, subscribe here and never miss a beat.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5bf4751d041e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Where Voice Tech Is Headed — Takeaways from AutoTech & Key Industry Trends]]></title>
            <link>https://medium.com/@kardome/where-voice-tech-is-headed-takeaways-from-autotech-key-industry-trends-9c9884216649?source=rss-45676c928070------2</link>
            <guid isPermaLink="false">https://medium.com/p/9c9884216649</guid>
            <category><![CDATA[car-vui]]></category>
            <category><![CDATA[voice-ai]]></category>
            <category><![CDATA[mobility]]></category>
            <category><![CDATA[speech-recognition]]></category>
            <category><![CDATA[consumer-electronics]]></category>
            <dc:creator><![CDATA[Kardome Technology LTD]]></dc:creator>
            <pubDate>Wed, 18 Jun 2025 13:55:08 GMT</pubDate>
            <atom:updated>2025-06-18T13:55:08.456Z</atom:updated>
            <content:encoded><![CDATA[<h3>Where Voice Tech Is Headed — Takeaways from AutoTech &amp; Key Industry Trends</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Pp_0yDK3FBgpeYqDr_N_nA.png" /><figcaption><em>The evolution of voice technology is moving beyond the ideal scenarios of quiet living rooms and personal assistants. Today, the real challenges and advancements occur in complex, noisy environments. This newsletter provides an overview of key developments demonstrated at AutoTech and broader trends shaping the voice AI landscape.</em></figcaption></figure><h3>Observations from AutoTech</h3><p>When we talk about the future of voice technology, we often imagine quiet living rooms and personal assistants that never miss a command. However, the real proving ground for voice AI is much tougher, filled with engine rumble, overlapping voices, and unpredictable road noise.</p><p>That’s why our demo at AutoTech this year marked a decisive moment for Kardome. We gave automotive OEMs an immersive demo of Kardome’s Spatial Hearing AI under one of the harshest acoustic conditions: simulated peak-level road noise.</p><p>Despite an intentionally chaotic soundscape, our voice UI software consistently isolated, identified, and understood individual speakers no matter where they were seated, feeding clean signals to the car’s speech recognition engine.</p><p>The demo impressed engineers and UX leaders alike, reinforcing that robust spatial hearing is no longer optional — it’s essential for intent-based, in-car voice UI that works in the real world, for everyone in the cabin.</p><p>We also showcased our collaboration with SoundHound AI by integrating their conversational <a href="https://www.soundhound.com/voice-ai-solutions/automotive">Automotive Chat AI feature</a>. Together, the system hears like a human and responds like one, identifying who’s speaking, filtering out distractions, understanding what was said, and delivering accurate, real-time responses, whether it’s the driver or someone in the third row.<a href="https://www.kardome.com/blog-posts/kardome-soundhound-ai-team-up-transform-in-car-voice-ui-multi-zone-conversational-ai"> &gt; Read more about this collaboration</a>.</p><h3>Key Partnerships and Integrations</h3><ul><li><strong>LG + Kardome</strong>: LG Electronics selected Kardome’s Spatial Hearing AI to power its next-generation smart TVs, with future expansion into smart appliances, audio products, and vehicles. This integration brings human-like hearing to consumer electronics, enabling seamless, accurate voice UIs even in noisy, multi-speaker environments. &gt; <a href="https://www.forbes.com/sites/johnarcher/2025/05/13/lg-oled-and-lcd-tvs-are-about-to-get-easier-to-use/">Read more on Forbes Innovation</a></li><li><strong>Kardome + NVIDIA</strong>: Kardome’s tech is now integrated with NVIDIA’s DRIVE AGX platform, delivering precise voice isolation, speaker identification, and improved speech recognition, enhancing the in-cabin voice experience for next-gen vehicles. &gt; <a href="https://www.kardome.com/blog-posts/kardome-mobility-nvidia-agx-spatial-hearing-ai-voice-recognition">Learn more</a></li></ul><h3>Voice UI and Safety Standards</h3><p>As vehicles become more feature-rich, the risk of driver distraction grows. In a recent article for <em>EE Times Europe</em>, Kardome CEO Dani Cherkassky explores how voice technology is moving beyond convenience to become a safety-critical solution. With European regulators increasingly focused on reducing cognitive load and minimising eyes-off-road time, in-cabin voice interfaces are emerging as a key enabler of safer driving, helping drivers stay focused, react faster, and comply with evolving safety standards. &gt; <a href="https://47487960.fs1.hubspotusercontent-na1.net/hubfs/47487960/News/EE%20Times%20Europe%20Dani%20on%20Voice%20UI.pdf">Read more</a></p><h3>Industry Trends and Highlights:</h3><ul><li><strong>Apple at WWDC 2025:</strong> Apple unveiled iOS 26 at WWDC, introducing the sleek “Liquid Glass” design and powerful on-device AI features like Live Translation, Hold Assist, and enhanced Shortcuts automation. While impressive, Apple delayed its Siri overhaul until 2026 due to quality concerns.<strong><br>Why it matters</strong>: Apple’s bet on private, on-device voice AI is a wake-up call to OEMs — users expect voice systems to be fast, local, and secure. However, even Apple’s challenges show how difficult it is to deliver consistent, context-aware voice UX.<br>Source → <a href="https://www.theverge.com/news/685106/apple-ai-siri-delay-craig-federighi-greg-joswiak-wwdc-2025">The Verge</a></li><li><strong>Google Gemini 2.5:</strong> Gemini 2.5 introduced live, interactive voice with emotion recognition, multi-voice TTS, and real-time audio APIs. These tools offer developers granular control over voice personality and interaction.<br>‍<strong>Why it matters</strong>: This is a blueprint for OEMs looking to create branded voice assistants that feel distinctly “theirs” — and more human.<br>Source → <a href="https://blog.google/technology/google-deepmind/google-gemini-updates-io-2025/?utm_source=chatgpt.com">blog.google</a></li><li><strong>Voice AI Startup Surge:</strong> In 2024, funding for voice AI startups surged 8× to $2.1 billion, according to PYMNTS. Startups like ElevenLabs and OpenAI are delivering lifelike AI voice agents capable of handling customer support, sales, and service at scale.<br>‍<strong>Why it matters</strong>: The democratization of high-quality voice tech opens doors for OEMs of all sizes to build custom, reliable voice systems into their products.<br>Source → <a href="https://blog.google/technology/google-deepmind/google-gemini-updates-io-2025/?utm_source=chatgpt.com">PYMNTS</a></li></ul><h3>Future Outlook for Voice AI</h3><ul><li><strong>Real-World Ready</strong>: Spatial hearing and understanding are no longer optional — it’s foundational for voice tech that works where we live, drive, and move.</li><li><strong>Privacy-First</strong>: On-device AI is critical for fast, secure, and trusted in-vehicle interactions.</li><li><strong>Personalization Wins</strong>: Customizable voices and UX are becoming key to brand differentiation.</li><li><strong>Startup-Driven Innovation</strong>: With funding booming, expect more rapid, affordable voice integration across industries.</li></ul><h3>What’s Your Take?</h3><p>Which voice-AI trend will define customer experience in 2025 and beyond — whether it’s the ear in the car, the assistant in the home, or the agent in your pocket? <a href="https://www.kardome.com/contact-us">Send us a message</a>, or <a href="https://www.kardome.com/subscribe-voice-tech-newsletter">subscribe here</a> to stay updated on the latest in spatial hearing and intelligent voice UI.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9c9884216649" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>