Conversational Artificial Intelligence (AI) is no longer science fiction, but an increasingly mainstream activity, with consumers interacting daily in their homes, workplaces and on the go. Usually known as bots, chatbots, or virtual assistants, conversational AI exists in a crowded and confusing enterprise market, with many bot versions that may not communicate with each other.
To help clear up some of this confusion, IBM Switzerland introduced the Watson Chatbot Challenge, an inter-university course involving Fachhochschule Nordwestschweiz (FHNW), Hochschule Luzern (HSLU), University St. Gallen (HSG) and Hochschule Zürich (ZHAW). The challenge, with 14 teams and a total of 69 students participating, is sponsored by industry partners, including CREALOGIX.
The objective of the challenge is for teams of students to acquire knowledge in conversational AI by solving real use cases based on IBM’s AI technology, Watson Assistant, a virtual assistant solution that allows users to interact with business systems using natural human language.
CREALOGIX use case
One of the use cases was set by CREALOGIX. With banks offering many different services and products, the challenge for new or existing customers is not to be overwhelmed by the amount of information and the structure of the sources, i.e. various websites with linked documents. Often bank agents need to answer questions regarding the conditions and prices of the products/services on the phone, assuming the customer has enough patience waiting in the queue. An integrated virtual assistant can be a big help in process optimisation, saving costs and improving the user experience.
Three teams of students took the challenge to build such an assistant. In the first design phase they looked at the different user stories – e.g. enquiring about the cost of an international bank transfer – and defined the words and synonyms that are key to identifying customer intent. They then created corresponding dialogue models that define the optimal flow of individual conversations, leading to the best outcome.
A very important part of the challenge was to feed into the system the specific information available on websites and FAQs and perform an extensive testing and training phase.
Positive feedback
At CREALOGIX, we were very excited about setting the use case challenge. The students were able to create and deliver three unique prototype chatbots with minimal time and prior knowledge, all excelling in their own way. From the prototypes CREALOGIX developed a service framework for customers in financial services that can be adapted to their individual needs. The solution incorporates a range of use cases, that can be offered on a personalised basis. Customers can easily and conveniently request information on products, conditions or fees themselves online via the virtual assistant, which can be trained in multiple languages and is scalable enough to meet increasing customer demands.
The feedback from the ZHAW team has been very positive:
"We are very grateful to IBM and CREALOGIX for the opportunity to apply the knowledge about AI that we gained in our lectures. In mapping the individual bank’s products against customer questions in the conversation, the virtual assistant was able to look for a possible answer on the website, or in the bank’s documents."
IBM’s Lars Mallien, who is running the challenge, concluded:
"It has been a great success. The students have learnt the concepts and logics of building a chatbot, and moreover I am astonished by the variety and results of the presented virtual assistants. In the use case presented by CREALOGIX the students have shown that there are various ways to solve a business problem in a very distinctive way."
Please find further information in the related press release.
