Communications solution provider Avaya Holdings Corp is planning to roll out a social networking platform for chatbots next year which will enable bots to engage with each other and find answers to customers' questions that they have not been trained to answer.

The company has already applied for the technology patent and is expecting to release it for commercial use by the first quarter of the next year, Ahmed Helmy, Avaya's Solutions Architects Director for the Asia-Pacific (APAC) and Europe, Middle East and Africa (EMEA) regions, told IANS.

The US-based business communications leader showcased the platform at the Gitex Technology Week being held here in Dubai from October 14 to 18.

Drawing parallels from traditional social media, the technology provides a structured platform for bots to engage each other in a secure and controlled manner with the intent of extending the expertise and effectiveness of each individual chatbot.

Helmy said the biggest challenge for chatbots to service customer needs was to answer questions that don't fall entirely in their domain.

While acknowledging the growing customer acceptance, a key shortcoming of enterprise chatbots is that they are domain-specific and can respond only to a relatively narrow set of dialogues.

This has limited their ability to fully and efficiently service customers' requests when they face questions they haven't been trained to answer.

"But what if we could enable chatbots from different domains and industries to collaborate and exchange information via a highly regulated platform, and find answers from other chatbots, not just humans?

"We could, in effect, securely link enterprises, significantly expanding the efficacy and expertise of any single bot, thereby enabling organisations to deliver an exceptional level of customer service, reducing the need for human intervention and enabling deeper self-service, as well as reducing the need for supervised learning," Helmy said.

Giving an example, he elaborated: "Like a chatbot for a restaurant aggregator may be able to recommend restaurants based on the customer's preferences. But if the customer wants to know if there is any special offer in a particular restaurant on his credit card, the chatbot may not be able to answer.

"However, if the same bot can refer this question to the said bank's chatbot, then it may be able to service the customer's need even though it wasn't trained to do so. This is the idea behind this social network for chatbots," Helmy explained.

A 2018 global Avaya survey of 8,000 consumers found that 80 per cent of people expect an immediate response from their banks, hospitals, hotels and even their governments, highlighting the need for efficient and 'always-on' customer service.

"With automation, organisations can overcome their human resource limitations and meet these customer expectations by delivering seamless, intuitive and intelligent experiences across all touch points. Bots are rapidly becoming one of the most powerful means to positively impact customer service, second only to face-to-face interaction," said Laurent Philonenko, Senior Vice President (Innovation) at Avaya.

Once deployed, Avaya's platform would allow any enterprise to register its chatbot with a unique social profile and 'friend' other member bots from different domains and industries.

It will also allow chatbots to rate each other, and store confidence metrics based on the quality of information received, and feedback from end customers, leading to constant improvements in each bot's quality and speed of customer service, a company spokesperson said.

"This is a huge step forward in addressing the information and service bottlenecks of chatbot systems," said Helmy. "The social platform model also means that Avaya's customers can increase the value of their chatbot solutions without having to engage in lengthy and costly data curation or warehousing projects."

Read more news:

India a Key Market in Asia Pacific For Expansion of Data Centres, JLL Report

Velocity MR Study shows Indians holiday now more than ever before