DT turns to biometrics for authentication and fraud detection

Deutsche Telekom has selected Nuance’s biometric technology to help the team offer customers simplified authentication processes when calling the customer service hotline.

Once the inevitable bugs have been worked out of the system, customers will be able to speak their requests naturally instead of navigating a complex phone menu, while the sound of their voice can also be used to confirm their identity. It is still early days, but the cumbersome process of typing in long account numbers and trying to remember complex passwords with a capital letter, number, punctuation mark and human sacrifice, could be a thing of the past.

“We’re proud to be leading the way as the first German telecommunications provider to deploy voice biometrics on our service hotlines and making this advanced technology available to our customers,” said Ferri Abolhassan, Managing Director Service at Telekom Deutschland. “We can identify our customers quite simply and quickly by the sound of their voices and there will be no more time wasted searching for contract numbers. The procedure is one of the most secure available.”

As far as stereotypes go, the DT management team must be giddy imagining how efficiently calls will be directed around the customer service centre.

“It is no secret that consumers today have higher expectations and demands for the type of service they receive from the companies they do business with,” said Robert Weideman, GM of Nuance’s Enterprise Division. “Our conversational AI solutions enable Deutsche Telekom to power more natural customer service conversations and deliver individuals the help they need quickly and securely.”

Although there will of course be sceptical customers who will rigidly stick to the old cumbersome ways, those who embrace the technology will simply have to say ‘Bei der Telekom ist meine Stimme mein Passwort’ (which means ‘At Telekom my voice is my password’ in English) to identify themselves. It’s simple and efficient, the way customer services should be.

Voice biometrics work by digitizing a profile of a person’s speech to produce a stored model voice print. Each spoken work is reduces to segments composed of several dominant frequencies called formants, with each segment subsequently having several tones that can be captured in a digital format. The tones collectively identify the speaker’s unique voice print, which the Nuance technology will allocate to a specific customer.

Aside from reducing call times and improving the experience when attempting to find the right department, the technology can also help prevent fraud. The voice print itself is similar to a finger print in that they are unique to individuals. While no system is ever going to be 100% fool proof, the voice print certainly does sound more secure than security questions, the answers to which can be worked out by effectively profiling a potential victim.

AT&T is a company which has faced complications in this area recently. Michael Terpin is suing AT&T for $224 million following the theft of $23.8 million in cryptocurrency tokens which were transferred out of his account following a SIM swap con. To do this, the fraudsters would have had to gain access to Terpin’s account by answering the security questions. Getting around these questions has become somewhat easier in recent years, as more user information has become available through social media.

In theory, fraudsters could research what the answers would be by look at Facebook interests, employment history on LinkedIn or holiday snaps on Instagram. Security questions are hardly the most creative and this is an area most companies should look at addressing. Using a voice print to authenticate the identity of customers should remove a substantial amount of the risk associated with such con jobs.

While this is an interesting and useful development at DT, it is slightly behind on the times. Voice biometrics have been used by numerous organizations, with Royal Bank of Canada, Santander, TalkTalk, and Vodafone Turkey among those who have each enrolled more than 1 million customer voiceprints after implementation. That said, it is a nice development.

Microsoft continues Nadella’s quest for conversational AI

It is a strategy which we have been hearing a lot about over the last couple of months, and the inclusion of Semantic Machines adds further ammunition to the Microsoft AI garrison.

This acquisition is a perfect example of where CEO Satya Nadella wants to take the Microsoft business. When Nadella inherited the Microsoft throne, the business was in a trough, relying on a decaying legacy business while carelessly investing to chase segments which had already boomed. The Nadella strategy was to get out in front of the crowd with a heavy emphasis on cloud computing in the early days, and now a more acute focus on artificial intelligence. Semantic Machines builds on this vision.

“We are excited to announce today that we have acquired Semantic Machines Inc., a Berkeley, California-based company that has developed a revolutionary new approach to building conversational AI,” said David Ku, CTO of Microsoft AI & Research. “Their work uses the power of machine learning to enable users to discover, access and interact with information and services in a much more natural way, and with significantly less effort.”

The idea of conversational AI is a simple one. Take AI interactions from limited and specific questions to conversation based outcomes. With more of a focus on conversational AI, offerings like Cortana, the Azure Bot Service and Microsoft Cognitive Services will be able to become more intuitive, taking actions based on context and more nuanced enquiries. It becomes a much more natural interaction and will ultimately drive the normalization of the technology.

This is where the Semantic Machines acquisition could prove to be a valuable one, as it builds on the Microsoft mission statement. The objective of Semantic Machines is to develop a ‘new, language-independent technology platform that goes beyond understanding commands to understanding conversations’. While this does sound like a blue-sky ambition, the task is separated into several areas including extracting semantic intent from interactions, deep-learning models to enhance the concept of context, speech recognition, speech synthesis and reinforcement learning.

The conversational AI strategy does look like it is becoming a useful one for the Microsoft team, especially in the telco space, where operators are trying to reverse years of negative customer service experience. While it is not reasonable to assume 100% of customers would welcome virtual customer service agents, progress with conversational AI will make the virtual agents more intuitive, broadening acceptance, while also making them more useful.

TIM is one telco which has grasped onto the Microsoft AI experience for its customer services, and we suspect it won’t be the last. For industries which have traditionally struggled in the customer services space, the telcos being top of this list, Microsoft’s conversational AI department is certainly one to keep an eye on.

Microsoft starts gathering pace in the telco AI world

Telecom Italia and Microsoft have expanded a partnership between the pair to develop new AI-based tools and services to improve customer service and experience.

The announcement follows a similar tie-up with Telefonica, which saw the telco use Microsoft’s Bot Framework to create its own digital assistant Aura. At TIM, Microsoft will help the telco expand its customer touchpoints with new apps and an interactive voice response tool, to help simplify customer relationship management.

“The present agreement is a step forward in DigiTIM’s strategy, strongly committed to providing digitization of all processes to dramatically enhance the digital experience for best in class customer engagement and to create an effective digital journey,” said TIM CEO Amos Genish. “Today we confirm once again our restless commitment to the execution of the Industrial Plan, of which Artificial Intelligence is a key pillar.”

DigiTIM is the focal point of the telcos strategic plan to slimming down company processes, improve digital experience and create new revenue streams. Announced last month, the aim will be to create a ‘digital journey’ for the customer, customising multichannel interactions and using artificial intelligence to create digitalised customer service management. The digitalisation of all processes is key here, and Microsoft seems to be taking a leadership position when it comes to AI in the telco world.

It is the big data dream, allowing TIM to have real-time and in-depth knowledge of the needs of every single customer. Through a new, single platform, TIM has high hopes in the convergence of fixed and mobile services, video, music and gaming content, to better monetize customers.

This is the grand plan, though TIM will certainly have some hurdles to negotiate over the next couple of weeks. For such a wide-scale, capital-heavy, digital transformation plan to come to fruition there needs to be cohesion above all else. Unfortunately, activist-investor Elliott is threatening the harmony at TIM, trying to force through its own agenda of asset stripping and short-term investor pay outs, in what can only be viewed as a share ‘pump and dump’ exercise.

Vodafone has shown an AI-orientated customer service approach can work, complaints against the telco are dropping pretty rapidly, though conflict in the board room will undermine any transformation journey. Digital transformation is complex and not cheap, for the DigiTIM strategy to work it will need the support and co-operation from every corner of the company. It is difficult to see that happening right now.

AI is starting to get seriously smart, perhaps worryingly so

We’ve seen a couple of AI shout-outs this week with Microsoft targeting telcos’ worryingly poor customer service record and European scientists potentially creating the building blocks for Skynet.

Starting in the Microsoft offices, the team has paired up with Amdocs to provide an intelligent solution to one of the biggest questions which has faced the telco space in years; why are we so poor at customer services?

In the Net Promoter annual report, the team put together a ranking of all the verticals in the US. At the top of the list were physical retail stores, and at the bottom of the list were internet service providers. Cable/Satellite providers were second to last, and cellular providers were 17 out of 23.

Clearly there is something going wrong, and the crack team of Microsoft and Amdocs aim to fix it. How you ask; by introducing chatbots for the telcos, as clearly humans can’t be allowed to have any interaction with their customers.

The SmartBot technology actually comes from Amdocs, but the intelligence is going to be from Microsoft’s artificial intelligence solution, Cortana. This is where the success of the idea will come from. The basic (and advanced) understanding of sentiment, intent, natural language and emotion which we as humans (well most of us anyway) do so easily will need to be replicated in an AI solution for this to be anywhere near a success. This is all Microsoft.

However, Amdocs has contributed something, the machine learning components of the bot. Amdocs claim, although most other AI companies also claim this, it’s intelligence platform can be used to build individual personal experiences for each customer.

But why start in the telco customer service arena? Well, the NPS scores are usually so low even if the technology doesn’t work, it’s not going to cause that much collateral damage…

However, despite there being an army of nay-sayers arguing against AI, Amdocs has said we’re getting more accustomed to the idea. Research from the team has shown:

  • Half of consumers think AI will become crucial to their lives within five years
  • 85% of businesses believe customers services will be performed by bots within five years
  • 37% forget they are conversing with a bot during some of these interactions
  • While 73% want human like traits in their bot to perhaps confuse the distinction further. Politeness was the most popular trait (64%), though a sense of humour was there for a few as well (29%). And we prefer female bots as well (35% vs. 15%)

“Artificial Intelligence has been domesticated and normalised in the minds of many consumers. Where once ‘AI’ meant robots and sci-fi, now it is has become accessible, is in people’s homes, and is helping them to complete everyday tasks,” said Jonathan Kaftzan, Head of Marketing at Amdocs Digital.

The normalization and acceptance of AI from the general public is of course crucial for any of these big ideas to work, but when should the line be drawn? If acceptance gives permission for scientists to push the boundaries, who is mature enough to say we have gone far enough, or will we find out too late…

Leaving Microsoft and Amdocs behind, a team of European scientist has been trying to figure out how to make robots smarter, more independent and self-healing. Hailing from Lausanne, Lisbon and Brussels, the scientists have written a paper called ‘Mergeable nervous systems for robots’ (you can find the full paper here), which puts out an idea which could be the beginning of the end.

The basic idea is to have a swarm of robots, all of which have their own CPUs, but can merge together to achieve more complex or physically taxing tasks. Once the robots merge, one is designated the ‘brain’ which control the rest. These robots can function on their own, but once merged, the power (both physical and intelligence) is increased through the basic idea of scale.

“Here, we present robots whose bodies and control systems can merge to form entirely new robots that retain full sensorimotor control,” the paper reads.

“Our control paradigm enables robots to exhibit properties that go beyond those of any existing machine or of any biological organism: the robots we present can merge to form larger bodies with a single centralized controller, split into separate bodies with independent controllers, and self-heal by removing or replacing malfunctioning body parts. This work takes us closer to robots that can autonomously change their size, form and function.”

Where this perhaps doesn’t become a good idea is the intelligence aspect. Right now robots are controlled by piece of software in the cloud. If things are going wrong, you break the connection, rendering the robot to a piece of metal. In this example, the scientists are proposing putting the intelligence in the machines, which can be scaled as the machines merge.

But it’s not just the intelligence that scales, it’s also the strength, the range of skills and the adaptability of the robots. Sounds neat right, or have we just started the beginning of the end of humanity.