Cyara has launched an integration of OpenAI’s ChatGPT into Cyara Botium, its automated chatbot quality assurance solution. The combination will allow customers to accelerate the development of chatbots and voicebots and improve their quality.
The key to this is how ChatGPT, using its large language models (LLMs), can better interpret the questions asked by customers. Previously chatbots only understood the language used within the documentation that generated its knowledge. With access to LLMs, it can now interpret a far wider set of language customers use.
The example given by ChatGPT was for a banking application where customers might ask:
- “check balance”
- “what’s my balance”
- “how much money do I have.”
The inference is that ChatGPT will understand the context of the question and interpret it accurately for the Botium app to respond appropriately. This challenge of bots understanding a question was highlighted in a recent Forrester report that revealed 61% of consumers believe that a chatbot is often unable to understand what they are asking for. This new integration should help to reduce that percentage.
How Chat-GPT help Botium
The integration enables Botium to generate a list of variant questions to help train the chatbot. The list can then be used to help train the chatbot. While this is one example given by Cyara, there are more. For example, asking ChatGPT a simple question such as: “What are the ten symptoms of a broken tap phrased as a question” gives a list of ten questions that could be used to check the chatbot understands and can give an appropriate answer.
Cyara argues that this can help
- Reduce the time taken to generate training and testing data
- Reduce the time and cost of development and testing
- Improves the quality of the chatbot with wider test criteria
It isn’t clear whether ChatGPT also leverages customer data to further enhance the test criteria used or is just the ChatGPT data. The Botium AI-powered Data and Test Generator are generally available now to customers.
Christoph Börner, Senior Director, Digital at Cyara, commented, “Cyara is leading the charge in delivering exceptional conversational AI experiences by recognizing the power of large language models (LLMs). This integration is another example of our commitment to delivering cutting-edge chatbot testing solutions to our customers.
“Cyara’s integration of GPT-3 for training and testing conversational AI has not only elevated the industry standard for delivering exceptional chatbot experiences but has also played a pivotal role in shaping the future of CX.”
Enterprise Times: What does this mean
There is a rush to leverage ChatGPT by SaaS vendors, often without real consideration for the solution’s potential benefits. Cyara has joined the bandwagon with an exciting use case. There is little doubt that ChatGPT can be used to enhance Botium. The big question is, by how much? Will Cyara be able to track the benefits Botium delivers to customers? Will customers be able to leverage the integration to the fullest extent?
One of the questions about ChatGPT that few are answering is how to use the solution to its fullest extent. Like Google, ChatGPT needs a prompt that requires a degree of thought, if not training to get the best answer possible. Hopefully, Cyara will deliver some insights on how customers can take the best advantage of the technology to deliver the benefits that are undoubtedly there.
