We’ve all had our good and bad experiences with call centers. Some people consider it a necessary evil, others would say it quite literally saved their life one day.
If you think about the type of conversations you have with call center agents, most of the time it involves some form of basic information that has to be gathered. Think about your name, licence plate number and estimated location if you’re calling for road assistance, or your customer or account number if you’re calling your internet provider because your modem stopped working. For a call-center agent, gathering this information is fully scripted and leaves little room for creativity. It’s often the dullest part of their job that gets in their way before they can actually get to work and provide the caller with a solution for his or her problem.
But what if we could automate this proces? What if we could develop a solution that can help call-center agents to focus on the part in their job that actually matters: helping someone solve a problem. Enter voicebots.
Call Center Automation is nothing new. IVR systems (Interactive Voice Response) have been around for years and we’ve seen speech-to-text technology as early as the 1950’s (Bell Laboratories’ ‘Audrey’ system could already recognize spoken digits). In more recent years, thanks to cloud computing, we’ve seen an explosion of speech-to-text software, making it available for the general public. However, in the call center world, there’s still very little intelligent automation, based on spoken voice, going on. We’ve seen examples from banks for instance, where you can use the dial pad of your phone to check on your account’s balance, but this doesn’t involve any voice communication from the customer whatsoever.
Today, we’re exploring the possibilities of merging this software with chatbots, to create some form of intelligent automation, based on spoken voice, within existing call center solutions. That last part is important. Investing in a completely new call center solution takes a big bite out of your company budget. Adding a software component within your existing environment however might be a more cost-effective solution.
If you look at this from a conceptual point of view, a voicebot is nothing more than a chatbot with a speech-to-text and text-to-speech component on top of it. Since we’re well experienced in building chatbots, exploring the domain of voicebots seems like the logical next step to take. We’re now involved in two pilot projects and looking at different alternatives for the SST & TTS components (Google’s Speech API, Microsoft Speech API, Nuance technology, MyForce technology, …). Both projects deal with a partial automation of the call and have a strong link with human call center agents of our customers.
We’ll be blogging a lot more about this topic and the advancements we make in both projects. If you’re interested in this topic, feel free to get in touch for more information.
This blogpost is inspired by our cooperation with Arinti.