Natural language dialogue systems

Virtual assistants, also known as chatbots, are an exciting area of AI which delivers huge cost savings for companies and a good user experience when developed correctly.

You may be familiar with Siri, Cortana or Alexa, which have recently become household names. All of these are examples of natural language dialogue systems, an area which has experienced exciting growth since Steve Jobs’ release of Siri in the iPhone 4S.

The manager Thomas Wood spent 5 years developing virtual assistants and have worked on the chatbots on the webpages of Shell, IKEA as well as internal chatbots for companies such as AT&T and HP.

The challenges of chatbots

Developing a chatbot involves several difficulties that do not appear in other areas of NLP. A successful dialogue system must

  • remember a dialogue context
  • remember referents such as he/she/it and resolve to items mentioned earlier in the dialogue
  • react appropriately to unexpected inputs
  • potentially deal with spoken text (speech recognition and speech synthesis)

Chatbots on Cloud

Today machine learning has accelerated the development of virtual assistants and there are a number of vendors offering out of the box software which allows you to develop your own chatbot without needing to write code. For example, cloud providers Google, Microsoft and Amazon all offer cloud based chatbot solutions, which are quite user friendly to work with but slightly more expensive than their lesser known competitors’ offerings.

If you would like to consult with us on a strategy for virtual assistants please get in contact.