This article originally appeared on IoT Evolution here
Voice may be the last frontier for cloud, or it may be the best.
While the industry in general has been focused on cloud computing since the turn of the century, voice continues to be the most popular form of communications, becoming even more popular with the advent of voice activated assistants, natural voice technologies, AI integrations and more.
In other words, voice is still the sexiest app on the planet.
Voice became IP n the 1990s, and while many of the world’s largest carriers, operators and service providers hid the fact the VoIP was driving more and more sessions for less and less expense, at the time (say the year 2000) dropped calls, jitter, and other issues plagued the user experience so “mum” was the word.
In the world of financial services, voice trading and other collaborations based on real time human voice transmissions could drive millions of dollars a minute, or a fraction of a minute. As the buy-side and sell-side talked things over before executing large trades, the market kept moving and there was nothing more immediate than an instant conference call via a “hoot and holler” or “ring down” service.
This technology created multi-billion-dollar value businesses like IPC Trading Systems which connects hundreds of banks over their trader voice platform, even as that company evolves to bring more and more high value digital services, like voice recording, market data, and more to their embedded base while attracting growth categories like hedge funds and alternative trading systems.
While voice applications like “voice mail” may have a bad rep compared to instant messaging and other channels – the fact is voice is growing, even as “landlines” are collapsing, and voice is growing “over-the-top” meaning voice sessions that happen over the Internet, and in ways we could not have imagined in 2000.
These digital conversations are being baked into workflow and business applications in very creative ways, and yet, they will not flourish with good old-fashioned quality of service. This is particularly important in industries like trading, emergency response, and other mission critical real time communications use cases.
Recently, GreenKey, an innovator in secure, quality voice delivered over IP within the context of fully secure private “extranets” joined the FINOS initiative to further grow multi-vendor support for “open voice collaboration.”
This is an important moment, as voice continues to join the open source community, which is become the rule rather than the exception.
FINOS (the Fintech Open Source Foundation) is a nonprofit foundation promoting open innovation in financial services seems to understand this and roped in GreenKey along with Cloud9 Technologies (already partners) to announce the contribution of the GreenKey SDK to the FINOS Voice Program.
GreenKey’s SDK is a JavaScript toolkit that connects third party applications to their backend technology and includes components for popular front-end frameworks to speed community development.
GreenKey’s open source SDK is a key building block that furthers the current work already underway in the FINOS Voice Program to define open voice standards for the development community and, according to the companies’ announcement, “embodies the collaborative spirit between developers and the finserv community to make this new voice SDK easily accessible to an open ecosystem of development teams.”
Tom Schady, GreenKey Chief Technology Officer, said: “All of GreenKey is moving towards open source - this first release is but one of many. We’re excited to bring voice communication, backed by our innovative real-time financial transcription engine, into any JavaScript application with the SDK. Our GreenKey backend does the heavy lifting in telephony and speech AI while our SDK lets you focus on your business and users. We’re also looking forward to contributing our voice interface and financial voice skills to the FINOS Voice Program.”
“We are proud to facilitate the connection between open source developers and the financial services community for the benefit of the entire financial ecosystem,” said Gabriele Columbro, FINOS’ Executive Director. “This contribution to the FINOS Voice Program is a perfect example of industry wide collaboration to build a truly open ecosystem based on common standards across end users and even competitive technology providers.”
Leo Papadopoulos, Program Management Committee Lead for the Voice Program and CTO of Cloud9 Technologies commented, “It’s gratifying to see new contributions from other leading firms like GreenKey to FINOS’ Voice Program. We look forward to working alongside other vanguards in this space to provide creative and effective solutions within financial services voice technology.”
Columbro said, “We believe that voice, like any other communication channel critical to the financial services workflow, is bound to become an open ecosystem,along the lines of what we have previously seen with text-based collaboration on Symphony. Leading vendors including IPC have an amazing strategic opportunity to join and play a leadership role in our Voice Program, leveraging their existing extensive network to attract and integrate with partners in this open ecosystem, with the net result of a better and more complete offering to their end users in financial services institutions. This is the reason with our Board we are actively recruiting leading vendors in the space, including but not limited to IPC, and we welcome them to our Community.”
Asked about the real value in future voice applications, including running AI analyses over voice sessions, Papadopoulos said, “Future value will be gained by the ability to run analytics and machine learning on the call metadata. Call metadata includes the call detail records as well as the transcriptions of the recordings. This will allow customers to better understand what business is occurring within in their voice communications and how to better optimize their business because of this understanding.” And while we didn’t contact IPC, a competitor of both GreenKey and Cloud9 about their take, given that they were not included in the announcement, we did pursue more detail about how "open voice" will change the way voice recording is conducted, including live recording, storage, retrieval and analysis of conversations.
This is becoming a pressing issue for firms, particularly given the onerous task of recording voice sessions over the last few decades. To quote one source who preferred to remain unnamed, “there are still people in the basement with headphones listening to recorded voice conversations to comply with regulations.”
Papadopoulos, a pioneer in the voice trading and voice recording space said, “Voice is already open. Google helped everyone with that when they open sourced WebRTC in 2011. The barrier to voice recording is standardizing on how call metadata and recordings are delivered to systems that want to consume that data. To that end FINOS has a created a ‘Call Metadata Standardization Working Group.’ If we all agree to put out data from our voice system in the same way, it will allow companies the ability to focus on the data instead of different integrations to get the data.”
Aha moment: so this is not just about open voice but potentially open voice recording, which companies like NICE are certainly looking into.
When asked about regulatory evolution, Columbro said, “Improvements and standardization in the way voice data is both stored and accessed means that voice data can be more readily available to help firms meet regulatory requirements. Being able to uniformly feed systems can help general AML (anti-money laundering) and supervisory control systems improve oversight and regulatory compliance as well as assist with parts of specific regulations such as MiFID II (transparency), Dodd Frank (call traceability and integrity) or the Senior managers Regime (demonstrate greater oversight).”
All voice data stored, whether transcribed or not, is subject to the same high level of security as other data, Columbro further explained.
Asked about the risks of opening voice transmissions and sharing voice records and its security vulnerabilities, Columbro said, “Whilst there’s still an ongoing debate on this topic, we believe open source and open standards deliver a superior level of long term security, especially as vibrant communities are able to address vulnerabilities with a greater transparency and typically in a much shorter timeframe than any single proprietary vendor. Additionally, if vendors are able to collaborate on common standards and components of their platform, that in turn reduces the overall attack surface, by simply consolidating and reducing the overall amount of code end users have to interact with in their voice-based workflow. Having said that, FINOS will focus on common tooling and standards, while it remains up to the individual voice providers to leverage the work products of our Community to deliver ever more secure commercial offerings to their customers.”
Even as the financial services industry and particularly trading is being reimagined (to put it lightly) by the advent of blockchain technologies which will have a massive impact on how trades are executed, cleared and otherwise recorded, high value voice technologies and automation through AI and “augmented” apps is just beginning.
Papadopoulos said “When Google open sourced WebRTC it allowed 1000's of HTML5/JavaScript develops to deliver voice/video/IM right from browser-based applications. Many companies have popped up in this space and are delivering Amazing products. It is too difficult to predict which market will be most influenced. What I can tell you is that WebRTC communications in the hands of the everyday developer and not just telecom experts. WebRTC open source has 'democratized' voice/video/IM development.”
Papadopoulos also opined on offerings like the Google Cloud Platform (GCP) and Amazon's Amazon Web Service (AWS), both of whom are offering amazing tools to enable voice command technology.
“They are making it easier and easier to integrate this into platforms. I think we will see these types of technologies move more and more into business applications instead of just using them to turn on the lights at home.”
In the future, more sophisticated software will be brought to bear, including voice biometrics.
“Security is added in layers,” Papadopoulos said, “Voice biometrics can be another security layer but should not be used alone. On the other hand, when transcribing voice where there are multiple parties talking, biometrics can help identify the talker and can be used to add even more meta-data to a transcribed call.”
If you think voice is over, think again. Voice is and always will be the best way to convey information in real time, given the many layers of meaning, including emotion, embedded in human voices. Voice is just getting restarted.