We are always working hard to roll-out new tools and features across our Iris Conversational AI platform, as well as to strengthen our existing product line to better serve our customers.


In the past, one of the biggest problems our clients have faced is the significant time spent handling repetitive questions across customer service channels. Responding to the same question over and over again wastes precious time for customer service agents - time that could instead be spent answering more nuanced or specific questions that truly require their attention.

In response, we’ve updated our Natural Language Knowledge Base App (NLU KB). In simple terms, the app differentiates between a repetitive question that can be solved with a single automated answer, and an action that requires a conversational flow and possible agent intervention. By enabling the building of a searchable library of articles, our NLU KB automatically handles repetitive questions by instantly connecting customers to the answer they require, moving them through a conversation, or directing them to an agent.

Conversational AI Customisable Confidence Threshold

As part of a series of updates, the NLU KB also supports a customisable confidence threshold, enabling our clients to set a confidence threshold that specifically suits their data and the range of questions customers ask. Setting a lower confidence threshold means the App will match more articles answering more customer questions, whilst setting a higher confidence threshold means the App will only respond if the customer question corresponds strictly to the article. This new feature is supported by a ‘Did you mean?’ function, where lower confidence articles are then also shared.

Finally, we have also expanded the metrics used in NLU KB analytics. We now include key phrase extraction, which enables invaluable insight into what people are actually searching for. This allows clients to see where gaps may exist in their articles library and where additional articles may need to be developed to address customer's questions.

We’ve seen demand for the NLU KB already, with almost every customer adopting the feature to help address the volume of repetitive questions their teams face. If you want to find out more about NLU KB and how it could support your team, don’t hesitate to get in touch with our Customer Engagement Lead, Roshni Roy here.


Start the conversation

Ready to see Iris in your business?


Talk to us about how to use Iris to engage, empower and support your customers.
Take your real-time conversations to the next level.