London Calling and It's Probably Using Twilio
Twilio acquires WebRTC technologies, and unveils Voice Insights at its London conference for developers.
This week Twilio, the CPaaS market leader, is holding its SIGNAL event in London, a follow up to its U.S.-based Signal event that took place in late May. You may recall a number of significant moves coming out of May's event -- including the super cool cellular IoT service it announced in partnership with T-Mobile -- and Twilio has carried that momentum across the pond and made two equally interesting announcements this time around.Open Buying Season: Advanced Video
The most notable news is that Twilio is announcing its first acquisition as a public company. Twilio has entered into a definitive agreement to acquire proprietary WebRTC media processing technologies built by the group that created the Kurento Open Source Project. While details on timing are sparse, when the deal does close, the Kurento team will become part of Twilio and work on advancing the company's video capabilities.
People familiar with Twilio know that they already offer a video API so Kurento might seem redundant upon first look, but Kurento addresses a much different market opportunity. Twilio's current API enables traditional "see you, see me" video in which one party can see another or a small group through a peer-to-peer session.
Kurento's Media Server is complimentary to Twilio's video solutions and includes large group communications, transcoding, recording, and advanced media processing. These services are focused on advanced video use cases that can expand the overall addressable market for video. In particular, developers will now be able to work with more complicated video fields like AR, IoT, and robotics, the press release states. Kurento will give developers the ability to analyze, transform, augment, record, store, playback and process video in ways not possible before without a lot of custom software and costly hardware.
For example, if a person were conducting a press interview using Kurento and wished to remain anonymous, the solution is capable of recognizing the person's face and can automatically blur the video. Another interesting use case is for virtual shopping. A shopper could try on a dress looking in a digital mirror and then tap the screen to switch colors or patterns.
Kurento also adds to Twilio's enterprise push (see related coverage, "Twilio: Priced for Enterprise Scale"). The recording and transcoding capabilities are important for regulated industries such as healthcare and financial services. Kurento can scale to very large meetings so businesses can drop video onto every workers desktop and evolve video from being primarily peer-to-peer based with limited appeal, to a resource that everyone can use in almost any situation you can dream up.Listen to My Voice
The other piece of news from SIGNAL London is the unveiling of Voice Insights, which is API accessible network and device data that enable developers to build applications with more consistent call quality. When the quality of a call is poor, it can be from a number of different factors, including network issues or an underpowered device. But how does one know why the problem is occurring? Developers have had next to no visibility into anything other than that the call isn't working properly and a user is complaining.
Twilio's APIs have been developed with instrumentation built into them. This enables Twilio to collect data and provide insights such as whether the user is having problems or there are issues with the device. Developers can access these insights through the API and then push a message to the user. For example, if a user starts a call and the headphones are not plugged in all the way, the developer could create a notification with that warning.
I think everyone has experienced some kind of voice issue that has been overly frustrating. The Twilio press release references that issues become exacerbated in geographies with unpredictable connectivity, but even I've had many issues trying to browse or make calls in downtown New York or San Francisco when the city is filled with people using their mobile devices.
There are many examples of how Voice Insights works but the use cases can be bucketed into the following three categories:
- Discover and respond to issues during a live call -- Developers can push pop-up notifications to warn of poor network coverage or even alert users that they are on mute.
- Identify and diagnose network and device problems -- The Twilio console includes a dashboard that displays network and device performance for every call and correlates the quality metrics with caller feedback. This can be used to quickly identify and resolve problems. ZK Research data shows that 90% of troubleshooting time is related to identifying from where the problem is emanating. Real-time visibility can significantly cut this time down.
- Report on WebRTC issues -- Developers are able to capture and analyze WebRTC experiences across a number of different factors. For instance, the performance of Android devices versus iOS can be compared between mobile operators, or developers could compare how WebRTC performs on Wi-Fi versus cellular networks. This data can be used to help businesses better plan and deploy WebRTC solutions. A retailer may choose to delay the rollout of a mobile application until the in-store Wi-Fi networks have been upgraded, for example.
Twilio's announcement follows last week's news from Vonage announcing its next-generation voice APIs. It's great to see so much activity in the CPaaS market as it's a faster, easier way to build communications-enabled applications. I'm expecting to see this market stay red hot, and to see it be well represented at Enterprise Connect Orlando 2017 this coming March.