Smart Voice Assistant Startups to Converse With

Remember when it was pretty easy to spot the crazy people on the streets because they were the ones talking to themselves? The introduction of cell phones, and then smartphones, made it much harder to differentiate the sane from the disturbed because suddenly everyone was walking down the sidewalk seemingly talking to thin air. Now, many of us spend hours (by some estimates, five hours per day) on our mobile and smart devices. And then the only thing we’re probably talking to is the device itself. If you use a smart voice assistant for conversational interactions, it’s likely one that was developed by Amazon, Google or Apple. But a company best known for its music identification app just raised a ton of money to challenge the big boys.


Click for company websiteNo, it’s not Shazam, which was bought by Apple last December for $400 million. SoundHound out of Silicon Valley on May 3 raised $100 million to bring its war chest up to $215 million. That raise was good enough to buy it a ticket to the Unicorn Club, where private membership requires a valuation of at least $1 billion. Founded in 2005, SoundHound had raised about $40 million through 2011 then went quiet for quite some time. Beginning last year, the money started pouring back in, with a $75 million Series D in January 2017, fronted by leading AI chip maker Nvidia (NASDAQ:NVDA). The $100 million corporate round includes names such as Chinese tech giants Tencent and Midea Group, along with major auto companies Daimler and Hyundai. Orange S.A. (NYSE:ORAN), a French telecom company, also participated in the latest round.

SoundHound is best known for developing an app (as seen in the video above) that helps you identify a song from a snippet of music. You can even hum a tune and the app will usually pick the right song and artist from millions of choices. It turns out the SoundHound app was just a way to keep the lights on while the company’s scientists and engineers developed what would become the startup’s flagship technologies over the course of 10 years. In 2015, SoundHound introduced its own smart voice assistant search app called Hound that can perform complex online searches and Houndify, an AI-powered conversational system.

Release the Hounds

It’s really Houndify where SoundHound hopes to make its mark and money, offering an agnostic alternative universe to the Amazon and Google smart voice assistant ecosystems. The idea is that developers and businesses can integrate an independent smart voice assistant interface into their products and services, along with a host of tools, knowledge graphs and domains. The latter are programs that provide a conversational interface on specific topics, such as the weather, stocks, sports, local businesses, flights, and hotels.

At the heart of the Houndify system are a speedy speech recognition system and sophisticated algorithms for natural language understanding, a subset of natural language processing. Natural language understanding enables machines to understand what people are saying without requiring humans to talk like robots. That’s not an easy thing to do. We use slang, incorrect grammar and all sorts of manner of speech that have led to early retirement of more than one English teacher. It’s also a valuable tool if you want machines to turn into compliance officers.

The Houndify platform is full of interesting features:

Credit: SoundHound

For example, Hyundai, which is a customer as well as an investor, will integrate the Houndify platform into what the car automaker is calling its “intelligent personal agent,” which allows drivers to use voice commands for different operations. The system is designed to be proactive, predicting a driver’s needs and responding accordingly. For instance, the assistant could provide a calendar reminder for an upcoming meeting and suggest a departure time to account for traffic conditions. Drivers would also be able to do all sorts of hands-free operations, from making phone calls and sending text messages to checking weather and searching for music. Finally, the smart voice assistant technology will be able to control various vehicle functions, such as adjusting the air-conditioning or locking the doors. The power of the connected car could even be extended to the home, ordering the smart home cocktail blender to make a double martini while you’re stuck in bumper-to-bumper traffic. Knight Rider’s KITT could never do that.

Running with the Pack

The smart voice assistant platform from SoundHound still stops well short of general artificial intelligence, when machines will be able to understand everything—even women—a moment in time referred to as the Singularity. SoundHound refers to its AI architecture as Collective AI, a nod to its crowdsourced underpinnings. The platform allows developers to contribute and collaborate on Houndify, providing functionality to all sorts of domains without needing to access or even understand them. Think of Collective AI as a browser with various extensions, or small software programs, that can work seamlessly together. The Houndify platform’s Collective AI architecture leverages knowledge and data from some pretty big sources, including Yelp, Uber, and Expedia, with a particular emphasis on domains related to the automotive industry.

The Herd Mentality

So it shouldn’t be too surprising to see Hyundai and Daimler helping bankroll the $100 million round. Hyundai has said it will start installing AI-powered smart voice assistants in vehicles by 2019. Earlier this year, Mercedes-Benz, owned by the German automotive corporation, showcased its own AI-powered voice assistant. The connection to Orange S.A. is also pretty obvious: The French telecom company is likely interested because AI-enabled smart voice assistants are increasingly moving onto mobile devices like smartphones, including Alexa, as Amazon continues to dominate the smart speaker market. And the China connection? If you’re going to go big these days, you need China. Tencent’s ubiquitous WeChat social media platform and its billion users is quite the prize. The big tech companies like Google are pretty well banned from operating in the country. At this point, it looks like China is SoundHound’s to lose, though the push by Chinese authorities to expand their country’s own AI dominance could always complicate things.

And Off to the Races

SoundHound isn’t the only AI-powered smart assistant startup to raise money this month. In fact, two other companies also took in some cash, though on a more modest scale than the $100-million mega-round.

Click for company websiteFounded just last year, Silicon Valley-based Suki got $15 million in a Series A on May 1 to bring total funding to $20 million for its digital assistant. The startup is developing an AI-powered smart voice assistant for doctors. Let’s face it: If you’re a proctologist with your arm elbow-deep into a patient, it’s a real pain to take notes at the same time. Created by former Flipkart and Salesforce executives, Suki claims its AI-enabled medical assistant is more than a transcription service through voice recognition technology. The Suki smart voice assistant uses machine learning to grow “smarter” as it interacts with a doctor. 

Credit: Suki

It can reportedly search and retrieve patient data, summarize patient and doctor interactions, order prescriptions and more. The company says physicians who use Suki have cut their time spent on notes by more than half. That’s a lot of golf.

Update 03/05/2020: Suki has raised $20 million in Series B funding to keep expanding its workforce and create new services such as voice-enabled patient billing. This brings the company’s total funding to $40 million to date. 

Click for company websiteSo, technically, Passage AI isn’t strictly developing an AI-powered smart voice assistant. Instead, the Silicon Valley startup has raised $10.3 million, including $7.3 million on May 2, for developing and marketing its AI business chatbot. However, some of the same AI tech applies, such as natural language processing and natural language understanding, so that the machine could hold an intelligent conversation, even with Flat Earthers. The company offers a chatbot-building tool that requires no coding, working more in the mode of a drag-and-drop model used to build some websites. Passage AI is emblematic of the wider trend toward handing over many office tasks to robots or AI bots and secretaries.


It’s starting to feel that artificial intelligence is becoming less of an emerging technology and more of a disruptive one, where we find ourselves interacting with machines every step of the way. Until we can talk to them with our minds, the next best thing is to say hello to the future. And we might just hear more from SoundHound in the not-too-distant future.


Leave a Reply

Your email address will not be published.