Why you should add voice and intelligence to your apps
|Richard Harris in Big Data Friday, May 19, 2017|
Voice-enabled AI and conversational intelligence provider talks with us about the future of their industry.
We sat down with Keyvan Mohajer of SoundHound to learn more about their mission to "Houndify" everything. Take a look:
ADM: Please briefly describe what Houndify and Hound are.
Mohajer: Hound is the world's most advanced virtual assistant and search app that lets you speak naturally to the connected world to get stuff done. It's the first consumer product to utilize the Speech-to-Meaning recognition technologies from SoundHound Inc.'s voice-enabling platform, Houndify. Hound aims to be the gateway to using all the technology on your phone - it's the voice activated path to surface information and drive action for everything from booking a hotel with Expedia to finding a great local restaurant.
Houndify is a platform that gives developers the tools to voice-enable apps, services, and devices. Houndify is independent and free from operating system constraints. It works across all platforms, including iOS, Android, Windows, Unix, Raspberry Pi, and others. It's available for wearable technology and is the platform that powers Hound. We believe you can't create a platform without also creating a product yourself that uses that platform.
ADM: When did you begin work on these products?
Mohajer: Houndify, the core technology on which the Hound app is built, was in stealth development for 10 years, starting from my Stanford dorm room. Because we've been at this for so long, we were able to release a voice product that surpasses anything else out there.
ADM: How is Hound different from digital assistants created by Apple, Google, Amazon, etc.?
Mohajer: One of Hound's biggest points of differentiation is that it's open and available everywhere, so whether you're an iOS or Android user, you'll be able to download the same Hound experience filled with all the great integrations from our partners, including Uber, Yelp, Expedia, AccuWeather, and many more.
We take a very unique approach to deep meaning understanding that combines machine learning algorithms, large data structures and program instruction to create a massive search space. It's a highly connected knowledge graph where we create a huge searchable space so that when a query comes in, we search everything at the same time and choose the highest probability interpretation, which is usually right.
Hound is powered by Houndify, providing all the technology ingredients necessary for voice and AI integration: the world's fastest speech recognition, the most sophisticated natural language understanding, easy to use developer tools, knowledge graphs, and a large and rapidly growing number of domains, or programs of information. Domains provide a natural and fully conversational interface on specific topics, without requiring the users to memorize and use the specific phrasing of hard coded commands or skills.
ADM: What advantages does Houndify offer developers and brands looking to add voice interfaces to their products?
Mohajer: Houndify aims to be a platform agnostic, one-stop destination for all the necessary technology components for voice search and voice control. This includes very fast and accurate large-scale speech recognition, powerful natural language understanding, and built-in knowledge domains.
The technologies encompassed in Houndify took years to develop, and are therefore not available to most companies. You don't need to be a scientist to use the Houndify API. It is meant to be easy to learn and easy to use for all developers.
Additionally, Houndify allows brands to stay in control of their data and brand. It is the only platform that has this benefit. For instance, once you adopt Amazon's Echo platform, your users become users of Amazon. Users will have to log into their Amazon account every time and have to call your product "Alexa." With Houndify, we don't have a hidden agenda to hijack products of our partners and users get to keep all the data that rightfully belongs to them.
ADM: What are the advantages for those who want to integrate their existing services into Houndify and Hound?
Mohajer: Currently, there are more than 20,000 companies using our Houndify platform to add voice and intelligence to millions of devices and apps, including Samsung for its ARTIK platform with millions of connected devices, NVIDIA for car dashboards, with more products and apps coming out this year and next. It's an entirely open platform that's not tied to a specific device manufacturer or larger platform company, so it works on any device or app.
Houndify allows rapid development cycles, inclusion of new domains, and the ability to scale faster than traditional voice technology solutions. You don't need a special device to use Houndify, you can use the technology on any device or product you choose. Houndify is teaching people how to start to using more complex requests to get what you want.
ADM: Who are some of the companies that you've partnered with for Hound and Houndify?
Mohajer: Some of the companies include AccuWeather, Expedia, Uber, Yelp, Rand McNally, Onkyo, Sharp, Xignite and Tanscorp. The company's momentum on the Houndify platform continues to grow at a rapid pace with more than 20,000 companies and developers registered, ranging from Fortune 500 companies to startups and independent developers.
CEO of SoundHound
ADM: How do you see the voice and IoT market evolving in the next five years? What about in 10 years?
Mohajer: At SoundHound Inc., we believe the more natural the interface is, the more likely people will adopt it.
When people speak to devices and the devices talk back with a natural response, that helps drive adoption and people won't feel awkward talking to a virtual assistant who can't understand them.
The current approach to Natural Language Understanding doesn't work well because the technology that is currently being used by other companies only detect entities and keywords within a sentence. When we started working on NLU, we asked ourselves what future would look like. We saw a world like the ones shown on Star Wars and Star Trek: a world where humans can speak to any device. Our goal is for Houndify to make a world a better place, more productive and more efficient.
ADM: Why are Speech-to-Meaning and Natural Language Understanding such key advantages for SoundHound Inc., and why are they crucial components to the future of AI assistants?
Mohajer: Voice interaction will eventually become the preferred means of communication with devices. Houndify is not only faster than the competition, but also more accurate, and we attribute that to our core tech which we call "Speech-to-Meaning." Where other voice apps and personal assistants do speech-to-text and then text-to-meaning with voice queries, which causes a lag in between the query being asked and the software providing an answer, Hound and Houndify do this all at once, drastically improving response time. Speech-to-Meaning also allows for our software to understand the context of a question and what a user is really asking for, as opposed to other voice assistants where nuances often get lost in translation.
For example, you can ask extremely long or complex queries, like, "OK Hound, please show me hotels in San Francisco that are pet friendly and have a pool and free WiFi and are under $300 a night." Moreover, the platform is always expanding and getting smarter, so our developers and the end users benefit from that.
ADM: How will AI impact the everyday consumer, and how are today's businesses leveraging the technology for them?
Mohajer: AI is impacting consumers' everyday lives in significant ways already, as it's now powering the backends of today's popular apps and services. AI is the reason why music fans love Spotify's "Discover Weekly" personalized playlists; it helps our email spam filters get better each day; it lets our photo apps automatically tag who's in our vacation photos; and it's why voice search and assistant apps like Hound can handle even the most complex of queries, such as, "Show me nearby cafes that are not Starbucks that have free WiFi and are open past 8:00 p.m." AI is quickly becoming ubiquitous part of our daily lives, and its role will only increase as time goes on.
ADM: Will typing become obsolete in the near future and voice commands be the new norm?
Mohajer: It's likely that most connected devices will have an intelligent voice interface; however, not all use cases will require or favor voice.
Learn the best ways to organize your app development projects, and keep code straight, clients happy, and breathe a easier through launches.
Write and run code every step of the way, using Android Studio to create apps that integrate with other apps, download and display pictures from the web, play sounds, and more. Each chapter and app has been designed and tested to provide the knowledge and experience you need to get started in Android development.
How to create a profitable, sustainable business developing and marketing mobile apps.