ChannelLife New Zealand logo
Story image

Kiwi company builds AI assistant with IBM Watson

A Kiwi company is working towards creating an artificial intelligence (AI) bot with IBM Watson Assistant to guide auditors through training and career options.

Replikr, an IBM partner, is working with global quality assurance organisation Exemplar Global on Emma, a new AI that represents the first phase of scalable, personalised advice to Exemplar Global's user base.

This user base includes thousands of auditors around the world navigating the complex options for professional training and development, a demand that Exemplar Global has found difficult to accommodate with only a small team, the company states.

Exemplar Global CEO Andrew Baines, based in Cambridge, New Zealand, says that as the organisation registers higher demand for certification services, Emma will enhance the customer experience for key industry players seeking information on professional development and training.

He says, “We want to make training and certification more accessible and attractive, especially to new generations of auditors, without adding to the overall cost of using our service.

"IBM Watson makes it possible for our relatively small organisation to scale up our services hugely using a digitalconciergewhile improving the overall experience for customers as they develop their careers.”

He says thatas Emma learns, she will in time become more integrated with Exemplar Global's CRM and Learning Management Systems, offering optimised recommendations for training including conferences.

Replikr managing director Dr Peter Catt says, “IBM Watson Assistant is a trusted technology that allows Exemplar Global to service a large customer base with a very capable virtual agent, potentially speaking to hundreds of people from a range of cultural backgrounds simultaneously.Importantly, IBM Watson will also allow us to easily access information from other systems to improve the Exemplar Global customer experience.”

The implementation at Exemplar Global comes as IBM announced new capabilities for IBM Watson Assistant to provide a higher degree of precision in natural language processing and foster greater trust in outcomes derived from AI predictions.

This includes FAQ extraction and a new intent classification model.

The FAQ Extraction uses a novel NLP technique from IBM Research to automate the extraction of Q&A pairs from FAQ documents.

Currently in beta form in IBM Watson Assistants search skill, it is planned as a new feature to help businesses keep virtual assistants up-to date with the latest answers and reduce time-consuming manual updates.

The new intent classification model is now available in IBM Watson Assistant. It is designed to more accurately understand an end-users goal or intent when engaging with a virtual assistant and to enable administrators to train the system faster.

The model provides more accurate results from less data compared to commercial systems, according to the company. This can help businesses go live with virtual assistants in a few days while achieving high accuracy.