'

SoundHound raises $75m from Nvidia, Samsung to scale voice-enabled AI tools

SoundHound is looking to become a serious AI player, having raised $75 million to scale its voice AI developer platform locally and internationally.

SoundHound has raised $75 million from a consortium of investors including Nvidia GPU Ventures and Samsung Catalyst Fund, bringing the total amount raised by the sound recognition technology company to $115 million.

The Santa Clara, California-based company will be using the capital to help accelerate the adoption of its artificial intelligence (AI) platform Houndify, as well as for expansion into international markets including Asia and Europe.

SoundHound's genesis goes back more than a decade, when its app was used to detect a song playing in the user's background.

With ambitions to become a serious voice-based AI player -- competing with the likes of Amazon and Google -- SoundHound then launched its own voice-powered virtual assistant called Hound for iOS and Android phones in March 2016.

The app was the culmination of a decade of research and development, Keyvan Mohajer, co-founder and CEO of SoundHound, said in a statement. The goal with Hound was to create a voice-based AI that can understand multiple data points within a query, allowing users to ask longer, more specific questions.

SoundHound expects that the need for voice control capabilities will explode alongside the number of devices connected to the internet.

"We are at the inflection point of our long-term vision that every product or service needs to have a smart voice-enabled interface, and consumers have increasingly high expectations for this requirement, beyond simple commands or skills," said Mohajer.

SoundHound's Houndify platform has been designed to provide developers with the tools needed for voice and AI integration so they don't have to build voice control capabilities from scratch, the company said.

The platform's "Collective AI" architecture enables developers to tap into existing knowledge domains without needing access to, or a full understanding of, the underlying libraries.

"This results in a global AI with comprehensive knowledge that is always learning, is crowdsourced to domain experts, and is larger than the sum of its parts," the company said in a statement.

Houndify currently provides access to knowledge and data from Yelp, Uber, and Expedia, as well as 100 other domains such as weather, stocks, sports, flights, and hotels.

Unlike Amazon and many other large companies with their own speech engines, SoundHound claims that its tools can be deployed without having to abide by a strict set of rules or give up any control of brand assets such as customer data.

Within the first 12 months of launching Houndify, its tools have been embedded into more than 500 distinct products from cars to consumer appliances, with more than 20,000 developers registering to use the platform. Nvidia, for example, has been using Houndify's speech recognition and natural language understanding technologies in cars.

SoundHound's other investors in the $75 million round include Nomura Holdings, Sompo Japan Nipponkoa Insurance, RSI Fund, Kleiner Perkins, SharesPost 100 Fund, MKaNN, Global Catalyst Partners, Walden Venture Capital, and TransLink Capital.

Conversational AI company MindMeld launched its own Deep-Domain Conversational AI platform in November 2016, claiming it to be the next stage in the evolution of conversational AI. The platform makes it possible for companies to create voice and chat assistants that can demonstrate knowledge and expertise around any custom content domain.

In December, conversational AI company Conversica also raised $34 million to scale its AI-powered sales assistant, which automatically engages with prospective customers using natural, two-way email conversations until the lead converts into an opportunity or chooses to opt out.