Whether one believes Artificial Intelligence, or “AI,” poses an existential threat to humanity or offers limitless potential for economic growth and individual utility, there’s no denying the technology continues to make remarkable advances year over year. We fall on the side of being excited about the great potential that AI has to offer for businesses of all types and sizes in this Fourth Industrial Revolution. Because of the huge amount of volume, velocity, and variety that the data is being generated, having the capability of analyze it, uncover hidden patterns and insights to make faster, better-informed decisions will make the difference in the long term. Automation with robotics and smart robot assistants are leading the way.
One of the latest examples of the continuing advancement of AI technology is known as conversational AI, which describes a set of technologies that enable devices and apps to communicate with humans using their natural languages.
That’s a lot of jargon in just a couple of sentences, so let’s take a minute to define some key terms relevant to this discussion:
- Artificial Intelligence “AI”: Computer systems able to perform tasks that have traditionally required human intelligence (i.e., speech recognition, pattern identification and anomaly detection, self-correction, etc.).
- Natural Language: A language that has developed naturally, over the course of time, through extended use, as opposed to a language intentionally created for a specific purpose. Consider Spanish or Cantonese as compared to Python, Java, or C#).
- Conversational AI: AI technologies that allow devices and apps to communicate using natural languages.
Conversational AI platforms, or “CAIP,” describe the applications or devices that enable humans to interact with conversational AI. Popular examples include virtual assistants like Siri or Alexa, chatbots or virtual assistants that triage calls to customer support centers.
Capabilities of Conversational AI
The fundamental capabilities of conversational AI are speech recognition and understanding natural language. Speech recognition involves simply being able to receive spoken instructions or other input, and to process that speech into recognizable sounds and then text. This process is an impressive feat by itself. But, unless a human speaker is speaking in C# or Python, the text generated from speech recognition is useless to a computer application without natural language recognition.
The ability to process natural language means that the computer can essentially interpret domain specific utterance provided in English, Spanish, Hindi, etc., in other words, identify the task or user intent and translate it into computer language capable of driving its functions and answer back to the user with an appropriate response. Natural language processing can manifest itself in a text-based chatbot or command entry, such as a user typing instructions on a keyboard in their native language; or, in conjunction with speech recognition, it can manifest itself as someone asking their virtual personal assistant what the weather is or telling the virtual customer assistant used by their bank to transfer money from their checking to their savings account.
Use Cases and Competitive Trends
Because of speech recognition and natural language processing, the capabilities of conversational AI are potentially limitless. Virtual personal assistants like Siri and Alexa are already widely used, as are virtual customer assistants as in our banking example above.
Gartner, in its review of the current and likely future state of the Conversational AI Platform (CAIP) market, suggests four likely competitive trends among providers of CAIP:
Multichannel/Multimodal, with a goal of providing broad applicability and a consistent experience across a variety of channels such as email, web, text, mobile, social/enterprise messaging, and Interactive Voice Response (IVR).
Prebuilt Vertical - or Domain-Specific Focus, such as developing a robust CAIP focused on HR management or telehealth, that can be deployed more or less out of the box for specific industries or business units.
Open and Agnostic platforms that can relatively easily leverage existing natural language understanding applications or platforms from such companies as Microsoft, Google, and Amazon to add more robust conversational AI technology on top of an existing foundation.
High-Customization / Low- Or No-Code platforms that allow users without extensive IT experience to use a suite of raw capabilities and mold them into customized use cases.
These approaches to market and attributes aren’t mutually exclusive. NITCO has just partnered with Kore.ai which is a great example of a company that performs exceptionally well across all four of the competitive trends discussed above. Kore.ai's Experience Optimization (XO) technology enables businesses to harness conversational interactions with greater efficiency and at reduced costs while ensuring the highest customer, employee, agent and partner experiences.
Understanding the Differences in Conversational AI
It is important to understand the differences between chatbot, intelligent virtual assistants and conversational AI. Gartner has evaluated several of the leading ready built Conversational AI Platforms and have articulated these differences in the following table. This identifies the level of complexity by characteristics, typical use cases and the appropriate technology differentiators. NITCO chose to partner with Kore.ai because it has been recognized by its peers as on the leading edge of emerging technology solutions. In particular, Kore.ai offers our clients End-to-end multimodal Conversational AI skills and Orchestration to solve business problems and to offer efficiencies to scale.
Imagine improving the customer and employee work experiences with exceptional ROI outcomes.