The term "chatbot" is being used a lot these days, and I have to confess that I wasn’t very excited the first time I heard what they were. Uh, okay, a text-based interface for chatting with a machine? What on earth would that be used for? But as with anything new, all it takes is a couple of good real-world examples to illustrate how pervasive – and useful – chatbots are going to be for us all.
One of the first places that we’re really starting to see chatbots coming on line is in customer care. Significant operators such as Verizon and Telefonica have been working on chatbots that assist subscribers with questions about what’s in their rate plan, upgrading services, and helping to troubleshoot configuration problems.The bots use natural language,are available 24/7 and neverput the subscriber on hold. That’s efficient for both operator and customer, and personally, I like a company that is respectful of my time. Look for a lot more developments in this area, including our Nokia Service Management Platform (SMP). We’ve built SMP to interface with common consumer interfaces such as Facebook Messenger and Alexa, so that this is the kind of interaction we would see:
Subscriber: "Alexa, what is my guest Wi-Fi network password?"
Bot: "I can send that to you. Do you want me to send it via e-mail or text message?"
Subscriber: "By text message, please."
Bot: "Okay, a text message with your guest Wi-Fi network password has just been sent."
How useful is that!? The end user gets immediate resolution to a fiddly, yet common problem, and the operator is able to increase customer satisfaction with zero human time expended. Win-win!
Beyond the signficant advantages of bringing chatbots to customer care, even more bot power is revealed when they link not only to the customer data of one person, but to the recorded experiences across an entire community. A good example of what I mean here is MIKA, Nokia’s Digital Assistant for network operations staff. MIKA is a natural-language voice and text interface linked to our AVA Knowledge Library that our operations personnel can use for network troubleshooting and repair.
Here’s a sample conversational flow:
Engineer: "I have alarms 70359 and 3159 at eNodeB site #23."
Bot: "OK, we need to fix 3159 fast – it’s service affecting! Supplementary information field indicates excessive operating temperature."
Engineer: "I replaced the unit, but we still have alarm 70359, and I have amber LED."
Bot: "Alarm 70359 indicates an un-mounted disk. This fault has been seen before in other networks. Here is the resolution . . ."
Did you catch that? MIKA is not only able to respond with alarm prioritization information based on the realities for this individual network right here and now, but it’s also able to compare today’s situation to a whole service history of past events from this network and others, in order to identify the best solution for the current problem. In the customer care example, the chatbots are doing what a very well-trained human would be able to do, but in this MIKA network operations example, the chatbot is able to consolidate and draw upon all of the service data from across all of Nokia’s networks, which vastly exceeds the knowledge and experience that any one human being would be able to accrue.
The next step beyond this will be Machine Learning, in which information about the correctness of the solutions proposed by the chatbot will be fed back into it, so as to make it even more likely that the chatbot will give the very best solution in its very first response next time And then after that, we’re in full-blown Artificial Intelligence territory, in which our computers are informing us about what they detected and what they fixed, while we’re enjoying cocktails on the beach. Or not...anyway, you get the idea.
But a journey of a thousand miles starts with a single step. Chatbots are that first small step towards automation of routine service tasks for both customer care and network operations that will continue to free up a few minutes here and there in our busy lives both at work and at home. And like I said, I like a company that respects my time.
Share your thoughts on this topic by replying below – or join the Twitter discussion with @nokianetworks using #AI #CSPCX #automation