Throughout 2019, we focused our roadmap on bringing more AI into the platform to take advantage of the developments in that field. During the year, we’ve worked hard to develop and deliver a series of robust AI features which simplify things for our clients and dramatically change the way they use our platform.
In the first quarter of 2019, we released our Long Question Engine – a feature which narrows down long input and pulls out the essence of it. This ensures that long input such as tickets, emails, and long questions can be recognised just as well as shorter input.
We’ve also released an autoAnswers feature which shows our continued efforts towards Natural Language Generation (NLG). The autoAnswers algorithm scans your website for information and automatically generates an answer to a question when there is not yet an available answer in the knowledge base. This means hours of work spared in collating and editing content for the chatbot.
But we haven’t stopped there. In recent months, we’ve taken one of the biggest steps in content automation that we’ve ever made. Our brand new, exciting feature – autoDialogs – integrates machine learning in the core of the engine in a safe way, with hardly any work to get it up and running. And best of all, the benefits are huge.
The autoDialogs feature can handle ambiguous questions based on specific information already available in your database, such as Q&A and Patterns. Typically, if an ambiguous question is not recognised, the engine will send back a generic response – “Sorry, I don’t understand your question’’. Instead of this generic response, autoDialogs provides end users with an automatically created Dialog containing a few options that best match the question, based on answers which are already available in the knowledge base. Gone are the days of having to manually build dialogs and maintain them. That’s sparing yourself of more than 50% of the work associated with building & maintaining a chatbot.
Of course, autoDialogs still allow you to have full control over the content shown to the end users. If the AI’s algorithm doesn’t provide you with what you had in mind, you can tweak the dialog as normal.
To ensure the feature generates dialogs which are in line with your content style guide, we offer the ability to adjust the autoDialogs Template. There is also the possibility to disable the entire feature. But that comes with a trade-off. Disable the feature – more manual content maintenance work, less machinery. Keep the feature enabled – less content maintenance work, more machinery.
Using AI to stay ahead of the market
As the market evolves, we’re constantly evolving too, and we’ve started seeing AI as a larger component of our platform. With that in mind, we’ve changed and simplified many things, which enable business users with no technical background to create a chatbot and have it go-live within as little as a day. It’s never been easier – add your FAQs, tag your patterns, go live, and watch as the AI takes effect.
“Prior to working with CX Company, we had tested various platforms, but we were surprised by how simple it was to create content and analyse data in CX Company’s DigitalCX platform.” – Inge van Vuuren, Marketeer at DELA, The Netherlands.
“Using DigitalCX has made it an easy and enjoyable project to work on. The drag & drop features are very user friendly and the dashboard has provided us with a good insight into what our customers look for on our website.” – Yasmin Inkeroe, Customer Experience Team Lead at Vivid Homes, UK.