A realistic view of what a Virtual Assistant can do for a business, in particular the comment that we aren’t trying to recreate a human experience (the technology for that is still some way off), but we can create an engaging, fun, useful experience that people want to use and can reduce the volume of agent-assistance required, sometimes significantly.
Nina has excellent natural language understanding capabilities to get to the ‘intent’ behind what someone says or types. For example, if the user says “I’m moving house” to a banking virtual assistant, the intent could be that they are interested in mortgage products, or home insurance, or how to change their contact details. Nina can understand this and conversationally ask clarifying questions to get to the real intent.
Where Virtual Assistants have more of a challenge is sourcing an answer, especially as the answer may not reside on a single domain. Traditionally this has been done by indexing content or keyword spotting or some other mechanistic process that rarely provides a useful answer (or maybe it provides 2,157 possible answers hopefully ranked in order of relevance – hardly a good user experience). With Nina we can take several approaches to this, creating bespoke content that is tailored to a Virtual Assistant conversation, linking to existing content in a knowledgebase of FAQ, or searching more widely if Nina isn’t confident she has a good answer. It’s not perfect but it is very good, and done well can offer a really good user experience within the scope of a companies business.