What are Virtual Assistants?

As databases multiply in size and volume, it becomes more strenuous to effectively manage and satisfy the demands of key stakeholders such as enterprises, employees or external partners.

 

With the digitization of business and the development of online services, the term "virtual agent" is catching on in popularity. However, it’s important to fully understand this term, as it can be misleading. 

 

From a technological standpoint, a virtual agent refers to a software of artificial intelligence (AI) capable of exchanging information, interacting with a user by means of a text or speech in natural language (e.g. English) and performing specific tasks. They occur mainly in the scope of web-based applications, messaging applications or stand-alone applications.

 

In other words, these AI-driven tools redefined our communication norms as they behave like humans and are able to perform complex tasks. Virtual assistants support employees to fulfill their queries, have access to their payslips or even get information regarding their holidays schedule. They are capable of meeting customers' needs by helping them navigate a website, access specific information or get refunds after returning a product.

 

Virtual assistants and ChatBots are similar but different, which is why it is quite common to confuse them as both are conversational interfaces, differing significantly in their level of sophistication. The following table is a concrete way to visualize differences and similarities between a digital virtual assistant and a chatbot.
 

 ChatboxVirtual Assistant
Communication modeTextText & voice
Where the bot livesWithin a third-party messaging platform (e.g., WebChat, WhatsApp, MS Teams)Integrated into product platforms or operating systems, e.g., Apple iOS (Siri), Microsoft Windows (Cortana) or Amazon Echo (Alexa).
AbilitiesLimited specific purpose, focus on a targeted domain context, pre-learned set of input.Multi-context, multi-domain, extended machine learning capabilities, adaptable, tailored content, intuitive, emotional.
PurposeInformation acquisition, basic assistanceProactive support, task achievement, deep navigation, engagement
Technology
  • [LB1] [CS2] NLP (natural language processing)
  • Rule-based programs
  • NLU (natural language understanding)
  • NLP (natural language processing)
  • Software based service
  • Artificial intelligence (AI)
InterfaceConversational user interfaceChat-like interface / voice commands
ChannelsWebsites / messaging channels / mobile applicationsMobile phones / Laptops / TV /smart devices
Example: “I want to cancel an appointment”Looks for keyword “cancel appointment” > provide knowledge or article describing the process / steps of how to cancel an appointment.Will ask for the appointment reference > place request > asks if sufficient help was provided.