Guest post by Nikolas Kairinos
ChatBots and other voice-based apps, which are pre-programmed to field a range of questions from end-users, whether customers or employees, have been around for years.
More often than not, these systems have come up short when it comes to delivering a first-class user experience (UX) – missing the mark with more complex queries, or failing to understand some all-important context.
That said, recent advances in Conversational AI (CAI) look poised to change this state of affairs. Beyond the realm of home assistants like Alexa and Siri, recent funding rounds have pointed towards a clear demand to bring these platforms into new arenas – Brain Technologies, for example, recently secured a staggering $50 million in funding for its own natural language processing (NLP) powered search engine.
Even further still, research predicts that the global Conversational AI market will grow from $4.8 billion in 2020, to around $13.9 billion by 2025 – figures which should leave no doubt as to the vast potential that these technologies hold.
But there are still many misconceptions to dispel when it comes to distinguishing Conversational AI from ChatBot technologies, which deliver an entirely different UX. ChatBots have historically been used to provide basic and transactional customer support, and while the age of Conversational AI is unlikely to displace any of this technology, new platforms will be able to support a much greater range of business requirements.
Although these capabilities are a far cry from those of older interactive voice response (IVR) systems, some organizations might take some extra assurance when it comes to embracing these technologies. So, what’s the difference?
At face value, ChatBots and CAI appear to perform the same functions. However, this is not the case. Although older ChatBots are typically built to replace commonplace human interactions, with the ability to engage in basic conversation and even prompt users if they have not engaged for a while. Architecturally, however, these systems are built very differently from CAI platforms.
CAI systems do not require this level of upfront manual curation
Simply put, ChatBots work with a more rigid set of parameters – as well as being keyword-driven, they have a more simplistic understanding of semantics, and require a lot of manual curation. Namely, humans must manually program sets of keywords and synonyms that will trigger predefined responses. The result of this is that they are often not suitable for interactions that fall outside of the domain of these discrete ‘keywords’.
Say, for example, you have a complex offering such as a software product, and your end-users are digital marketers. Beyond offering basic information about the product to employees, users would be hard pressed to obtain concrete answers when asking more complex questions of an IVR system.
If the system were to break, for instance, they would need access to a support knowledge base to overcome the problem. At this point, you would likely be asking the ChatBot to do too much semantically, and the architecture will falter – leaving it unable to present solutions to more open-ended or unexpected queries.
CAI systems, on the other hand, do not require this level of upfront manual curation – as well as recognising a greater number of synonyms when end-users interact with the system, newer technologies that utilize neural networks will have a better grasp of language and semantics more generally.
Consequently, these systems are capable of offering less structured support, and allowing more conversational back and forth with the user. The use of sophisticated algorithms means that these platforms can remember previous interactions, ask follow-up questions unprompted, and even grasp the context of a conversation without obvious cues. This ultimately makes for a more seamless user experience.
What do businesses stand to gain from conversational AI?
As I have already noted, businesses stand to benefit from less structured support, but the use-cases for CAI go beyond this. Particularly as many organizations have now shifted to hybrid models of working, more businesses will be facing up to the fact that their employees must have the necessary knowledge to perform their roles remotely. In this way, constant Zoom meetings may not suffice, and workers must have the ability to find the right information with some more independence.
NLP technologies will enable employees to ask questions on the fly, without having to worry about whether their query is understood by the system, or having to schedule a lengthy call with a line manager. As such, these technologies clearly have a lot to offer in the corporate training arena, particularly when it comes to offering consistently delivered, year-round autonomous learning to workforces.
With the ability to type or speak their questions or commands into a platform, businesses will be able to benefit from technologies that truly understand their staff. This new generation of conversational agents will understand the minutiae of everyday conversation including abbreviations, slang and even different dialects.
Those working in B2B sales in particular will have a lot to gain from these technologies. Given the ease of conversation with CAI technologies, individuals will be able to pick up information on the go – if they were delivering an important pitch to a prospective client, for example. All that a user would need to do to refresh their knowledge of the USPs of a specific product is ask the necessary questions.
Ultimately, businesses must shift their old perceptions of underwhelming ChatBots with limited use-cases, and look towards newer technologies that expand their horizons. They certainly won’t look back.
Nikolas Kairinos is CEO of Soffos, the world’s first AI-powered KnowledgeBot.
Related
AI: the intelligent future for small businesses
Why every business should move to an AI-first strategy
