Houndify slides into DT voice strategy – to enhance or replace Djingo?

SoundHound, a dominant player in automotive voice-enabled AI, appears to have found a new niche. Following news this week that the company has partnered with Deutsche Telekom (DT), it seems that the Houndify platform is being used to enhance existing voice assistants, working alongside market leaders, such as Amazon Alexa.

This is not a position in which SoundHound seemed to be aiming just a few years ago. Investment rounds appeared to be gearing the firm up to be serious competition against Google Assistant and Alexa. However, if played right, this new accessory-type role in the market could prove lucrative.

Houndify has come to prominence as a voice-enabled AI platform designed to equip businesses with voice intelligence, enabling product integration across a variety of industry verticals. However, the exact purpose of the partnership with DT is so far unspecified – although one likely possibility is to allow the German operator to further hone the Magenta smart speaker, which it released in September.

Equally, the recent partnership with DT could provide further business in the automotive industry. Houndify has already partnered with Kia, Honda, Mercedes Benz, and Hyundai, while DT has special relationships with BMW, Audi, and Volkswagen.

Magenta was one product to come from a joint effort by DT and Orange to produce a voice assistant specifically catered to European markets. After much delay, Orange launched Djingo in Spring of this year, while DT announced its Magenta speaker in September this year at the IFA show in Berlin.

Despite both telcos investing heavily in their own voice-enabled AI, both speakers were released with Amazon’s Alexa included as a second voice assistant. We noted at the time how Alexa’s inclusion was a red flag for the Djingo AI ecosystem, a sentiment which doesn’t quite ring true with the latest integration. If SoundHound’s calling at DT is to enhance smart speaker development, Houndify’s fate is clearly to slot into products already holding a leading voice assistant, rather than establishing itself as a fully-fledged, consumer-facing assistant.

But is this necessarily a bad thing? The market leaders, Alexa and Google Assistant, are so ubiquitous that accommodating a third, similar-sized voice assistant is farfetched.

Rather, Houndify’s platform should be recognized as a quality enabler with which companies can enhance existing voice assistants. The company’s website is loaded with neatly worded features and qualities which hope to set the platform head and shoulders above its competition.

Speech-to-Meaning response condenses the traditionally binary processes of speech recognition and language understanding into just one. Query Glue allows multiple questions to be processed by the AI simultaneously. Knowledge Graphs provide the AI with conversational contexts – such as a user’s location or previous queries – to allow natural exchanges.

Houndify is also made smarter by its collaborative intelligence base – coined Collective AI – which allows partners and developers to extend the functionality of existing knowledge domains. There are also a host of Private Beta services available on request, including Beam Forming, Echo Cancellation, Emotion Detection, Automatic Language Translation, and Face Recognition.

Such jargon can be explained as creating representations that encode aspects of what we would understand as meaning – such as basic sentiment, speaker-attributes and other fairly basic and well-defined components studied in the field of natural language processing. As for achieving “meaning” as we would know it, Faultline has it on good merit from experts in the voice technology sector that the market is a million miles away from cracking the meaning nut, despite what some companies want us to think.

A similar story of service enhancement played out following Houndify’s partnership with US internet radio firm Pandora in 2018. The company launched an in-app Voice Mode which was powered by Houndify, but this occurred 3 months after Alexa had been integrated into the app.

SoundHound has seen a couple of large investment rounds in the past few years. We reported on a $75 million funding round in 2017 to fuel development of Houndify. This was followed by a further $100 million raised in 2018 to drive international expansion of the platform into China and Europe – across industries such as automotive, IoT, enterprise, and consumer products. Spectators saw this as SoundHound gathering arms to take on Alexa and Google Assistant.

Houndify’s most recent partnership in the automotive industry was with Hyundai. In June, the companies collaborated on bringing the first voice-enabled, smart-connected SUV to the Indian market – the Hyundai Venue.

Aside from Houndify, SoundHound operates two other verticals. The app it is perhaps most famous for, SoundHound, provides song recognition. The company’s other app, Hound, is a simple voice assistant, allowing one to carry out tasks on their handset without touch.