The true capability of artificial intelligence is astounding, and we are just starting to scratch the surface. While we are confident that what is available to us, as opposed to governments and military contractors, pales in comparison, the awesomeness of this technology is incredible. As quantum computing rounds the corner and the human race’s ongoing development of neural networks continues to evolve, it is only a matter of time before all of society is fundamentally altered by this technology.
As early adopters of the various tools made available by companies like IBM and AWS, we have developed a series of services that we make available as SaaS offerings. These services are available via our API Library, and have been specifically engineered to enhance your operation. While we are not experts on Artificial Intelligence, we have developed several applications that are very specific and helpful to organizations that either employ contact center services, or operate contact centers directly.
Anyone who operates a contact center, five seats to 5,000 seats, understands that the human element of your staff is the most challenging obstacle to success on a daily basis. Sickness, attitude, training, morale, drugs, food, support services, general culture – all of it adds up to a guaranteed headache, every day of the week. Bishop was born from an early adoption of tools available from both IBM and AWS to create a virtualized agent with the goal of offloading as much of the conversational dynamic from humans to technology, as a consumer would stomach. As the advancement of text-to-speech and speech-recognition has grown over time, so has the functionality of Bishop.
The chat version of Bishop is obviously more refined as Bishop is not required to sort through various dialects and speech patterns with a chat interface. Bishop can handle, at current levels of complexity, roughly 70% of your organizations email, chat, and SMS conversation flow for customer service and even sales. This requires a significant amount of machine learning model training, but once complete, Bishop can handle the bulwark of your digital conversations without bathroom breaks, illness, or any of the other issues that a human workforce is faced with. Obviously, with the Pandemic of 2020, this is even more important than ever as your customers expect you to be fully operational despite a massive disruption to your human capital.
The voice version of Bishop is still very much in a an alpha refinement stage. At present the greatest challenge we have faced with voice is the complexity of model training for the average consumer. Many do not wish to interact with a “bot” so we have gone to great length to remove much of the machine specific voice elements and pacing. At present we feel comfortable stating that Bishop can handle about 18% of the voice traffic of a typical install, but requires significant model training and support to ensure the best possible experience for your consumers. We are in a constant state of training and evolution with Bishop and are excited to share information about Bishop with our clients, partners and vendors for the first time.
While Bishop may not be able to handle the lion’s share of your typical voice requirements, Bishop can handle incredibly effectively the collection of sensitive information – PCI, HIPAA, RESPA compliance is spectacular. Your agents simply transfer into a Bishop queue, where the conversation has been monitored and is continued for the collection of sensitive data, and then passed back to the agent. This ensures the security of sensitive data, especially important when your virtualized workforce is under even less secure protocols than normal.
Virtualized Quality Assurance
Thanks to our experience managing billions of phone conversations over the years, and the human teams necessary to ensure the quality of the calls – that agents are not making threats, saying negative things, making promises they cannot keep, lying to consumers, etc – we realized that as long as humans are involved, there is the opportunity for corruption, human error, and general laziness. Our solution to this challenge was the development of Viki, a virtualized call quality monitor that runs in real time, on top of every single phone call, email, chat, and SMS conversation.
By utilizing machine learning models that are specifically developed for regional dialects and languages, Viki can “listen” in to every conversation that your organization is having, and take action where appropriate. Viki can not only be trained for specific words and phrases, but can also pick up on intent signals. Once the training model is complete, Viki can take action by barging into a call and transferring to a tier 2 supervisor or senior skill set. Viki can send system alerts and notifications when an agent is using language that is inappropriate, and can also be used to complete a report card for every agent on every conversation. Storing these report cards in the Nerve Center, we are able to provide actionable intelligence for your organization to step counsel individuals, proactively solve for challenges with your customers, and to serve as a full time, non stop enforcer of your compliance goals.
If you want to learn more about Bishop or Viki, or have other ideas that you would like to speak with us about, please use the button below.