AI has changed everything for the UC and collaboration sectors. Conversational AI, specifically, will revolutionise the way that businesses internally collaborate and also interact with their customers.
There are myriad factors that businesses need to take on board when considering investing in conversational AI — from whether to seek multimodal conversational AI that can integrate multiple modalities beyond just text to what type of dialogue systems a business might want to implement within its conversational AI model.
With our latest Round Table subject, “Conversational AI”, we spoke with experts and executives from Vonage, Zoom, Sinch and Avaya about how conversational AI can enhance both UC in modern workplaces and the user experience, the challenges of implementing conversational AI for businesses, and how conversational AI might evolve in the future.
How does conversational AI enhance the capabilities of unified communications in modern workplaces?
Noam Mor, Senior Manager, AI Product at Vonage
Mor argued that businesses navigating modern workplaces are seeing the value in merging their UC and contact centre solutions with consumers expecting an easier transaction when doing business with a brand. Conversational AI helps address that need as it has “enhanced UC by meeting the needs of dispersed workforce collaboration while maintaining customer engagement,” Mor said.
“Conversational AI is a valuable tool for brands to deliver exceptional customer experiences,” Mor elaborated. “Contact centre agents rely on AI to anticipate customer questions and help resolve issues. Virtual assistants have relieved some burden on employees by helping resolve simple issues like rebooking flights. By automating responses and addressing requests through AI, businesses across industries can transform their processes to immediately connect with customers on the channels they prefer.”
Mor highlighted that generative AI could also improve the agent experience by predicting the next best actions, generating recommended responses and assisting with summaries while reducing training time. “This helps businesses harness more conversational power with lower investments,” he added.
Vijay Parthasarathy, Head of AI/ML at Zoom
Parthasarathy outlined Zoom’s bullish position on AI since the company began, highlighting its heavy investment in”fundamental technologies to help enable future AI experiences”. Parthasarathy cited Zoom’s conversational intelligence-focused products, such as Zoom Virtual Agent and Zoom IQ for Sales, as examples of them fulfilling that vision.
“We continue to embrace conversational AI as an integral part of our platform with a federated approach that prioritizes flexibility,” he added. “For example, in the recently introduced Zoom IQ Meeting summary feature, we leverage Zoom’s own LLM to simplify a complex dialogue so hosts can easily share summaries and, soon, action steps from meetings.”
“Team Chat Compose, also recently released, leverages OpenAI’s LLM to draft messages based on the context of a chat thread, in addition to changing message tone and rephrasing responses to customize text recommendations.”
Parthasarathy also said that Zoom Virtual Agent, Zoom’s intelligent conversational AI and chatbot solution, allows organizations to deliver efficient web and mobile support via its Zoom Contact Center solution to meet the expectations of customers for exceptional service while optimizing costs.
Joachim Jonkers, Conversational AI, Product Lead at Sinch
Jonkers believed there were two key areas where conversational AI could have an impact on UC — customers and employees working in the UC platform.
“For customers, conversational AI allows for personal conversations at scale,” Jonkers commented. “It does this by retrieving and saving information from external systems through integrations. Conversational AI systems are also available 24/7, which means customer service or any given inbound enquiry can be dealt with instantly, unrestricted by office hours. The user experience gains are significant.”
Jonkers also pointed out that conversational AI frees agents to focus on the questions that require a human response. He cited frequently asked questions such as ‘What is my account balance?’ or ‘When will my parcel be delivered?’ as requiring a response that can be easily retrieved and automated.
“In other nuanced ways,” Jonkers continued, “Conversational AI systems based on Large Language Models (LLMs) such as GPT can help write better campaign messages, help agents write better responses by making them friendlier or informal and can make a traditional one-way marketing campaign more interactive in all cases improving outcomes such as conversion rates.”
Neal McMahon, Regional Sales Leader, UCaaS and CCaaS, Avaya
McMahon underlined that AI is accelerating the blurring between customer and employee experiences. “Most people think of conversational AI as belonging in the contact centre, and certainly, most of the compelling AI improvements are in the customer service environment,” McMahon said. “But AI is now everywhere, and as we’ve seen recently with generative AI or prompt-based AI, we are getting into a new world of accelerated automation and orchestration.”
McMahon highlighted that conversational AI could not only continue to offer contextual background and guidance but help orchestrate that workflow. “With an experience platform, both UC and CC features are used to benefit the customer and the employees,” he added.
How does conversational AI improve the user experience?
Neal McMahon, Regional Sales Leader, UCaaS and CCaaS, Avaya
McMahon said that, as of 2023, there are no longer quarterly or bi-annual updates from large AI providers but an almost weekly improvement from many AI tools.
“Live call transcription on conference calls is an important AI feature that improves the unified communications user experience,” McMahon suggested. “This is the ability for AI to add a textual transcript of the conversation in real-time and is particularly useful for hard of hearing and/or anyone that has to be on mute/silent for any period of the call.”
“Another way that conversational AI improves the unified communications user experience is by cutting out background noise, so if you are on a call in a noisy location – and the point of UC is that you don’t have to be tied to an office where it’s quieter – everyone is heard clearly in real-time.”
McMahon also said that, at the other end of the scale, a UC solution that integrates with an experience platform as a single solution is customisable and can be tailored to solve whatever a business needs at any time. “With an open platform, hundreds of business apps can be integrated,” he added, “so if you need the features from conversational AI one season but not another, that is not a problem.”
Vijay Parthasarathy, Head of AI/ML at Zoom
Parthasarathy argued that conversational AI could help improve the user experience by providing critical insights that “sales leaders can use to develop their teams, enhance their customers’ experiences, and make more informed decisions in the future.” He highlighted Zoom IQ for Sales’s capacity to do just that.
Parthasarathy also mentioned Zoom’s recently announced strategic partnership with Anthropic to enable the flexibility customers want and significantly improve collaboration.
“The partnership allows Anthropic’s AI assistant, Claude, to be integrated with Zoom’s platform, starting with Contact Center,” Parthasarathy described. “With Claude guiding agents toward trustworthy resolutions and powering self-service for end-users, companies are able to take customer relationships to another level.”
Noam Mor, Senior Manager, AI Product at Vonage
Mor believed that conversational AI could enhance the user experience by eliminating phone trees, long wait times, and slow responses. “The technology can be implemented across a variety of channels, such as web, mobile, chatbots, or social, meeting customers where they are while incorporating context from previous interactions,” he said.
Mor also discussed how conversational AI-backed chatbots could extract data from a number of sources, including customer relationship management (CRM) platforms, customer profiles, and purchase history. “The AI can use this information to personalize interactions, anticipate needs, make recommendations, and remember the information for future interactions.”
“With conversational AI,” Mor continued, “a user can communicate with a brand via SMS or a voice assistant, and the AI is able to recognize the customer, remember the last interaction, and apply that context to the current one. It’s a seamless omnichannel experience personalized for each unique customer.”
What challenges and considerations should organizations be aware of when implementing conversational AI in their unified communications infrastructure?
Joachim Jonkers, Conversational AI, Product Lead at Sinch
Jonkers outlined five six considerations in considering the risks of conversational AI.
Firstly, data privacy and security, as conversational AI systems handle sensitive information. “It’s crucial to ensure that the platform has robust security measures in place to protect data at rest and in transit,” Jonkers said. “Encryption, access controls and secure storage are vital considerations.” Jonkers also pointed out the potential legal and regulatory requirements around AI and data: “Organisations must ensure compliance with applicable laws, such as data protection regulations or industry-specific guidelines.”
Thirdly, is that UC infrastructures often include various pieces of infrastructure, including phone systems, email, instant messaging and collaboration tools, which complicates the task of integrating conversational AI. “The process may require APIs, middleware or connectors,” Jonkers suggested.
Then there is user adoption and acceptance as employees, customers and stakeholders need to adapt to the technology. “User training and education programs should be considered to familiarise users with the capabilities, benefits and limitations of conversational AI,” Jonkers explained. “Clear communication about the purpose and expected outcomes can help in gaining user acceptance.”
The next key consideration is that organisations based in diverse regions or serving global customers have to factor in the multilingual and multicultural aspects of conversational AI. “The system should support multiple languages and have cultural sensitivity to cater to users from different backgrounds,” he said. “Localisation and translation requirements should be taken into account during implementation.”
Lastly, Jonkers warned that conversational AI systems could have “inadvertently inherent biases” based on the training data or biased behaviours learned through interactions. “Organisations should be vigilant about potential biases and take steps to minimise them. Regular audits and ongoing monitoring can help identify and address any bias in the system,” he added.
Neal McMahon, Regional Sales Leader, UCaaS and CCaaS, Avaya
McMahon raised the pertinent point that, like with any technology, AI will need to learn in order to improve: “It’s important to remember that AI tools can’t understand context or respond to the emotional state of employees. Additionally, the tools are based on data and text, not real-world experience.”
“It’s therefore critical these tools complement rather than replace the human element,” McMahon argued. “Many organizations are trying to understand how you control large language models, control the data that’s inside them and secure it, and then bring relevance to agents, self-service, analytics, or even internally to help build tools that can serve customers and employees better. There’s a large focus on delivering this in the right way.”
Noam Mor, Senior Manager, AI Product at Vonage
Mor also highlighted the potential growing pains of the technology, stressing the importance for the company to be on the same page in implementing conversational AI into its UC infrastructure. “Companies should first prioritize where their customers are and their preferred communications channels, as conversational AI is a technology that powers a company’s additional communication solutions beyond just traditional chatbots,” Mor said.
Mor also underlined AI’s daily evolution as a technology and warned companies to make sure they are keeping up with changing speech trends and customer needs. “This could mean new words to account for or even unscripted questions that an AI-powered bot can’t answer,” he added.
“While beneficial in contact centres, conversational AI can also be used to enhance IT support or improve a retail and e-commerce platform – companies should look across the entire enterprise to evaluate where conversational AI can best support the business.”
Vijay Parthasarathy, Head of AI/ML at Zoom
Parthasarathy recommended businesses pursue a federated approach to AI similar to Zoom’s because it “allows for flexibility and customization when incorporating multiple types of models, resulting in offerings that provide the most value for customers’ diverse needs”. “The full potential of use cases for conversational AI is not fully known, so being flexible in capabilities will be the best way to adapt,” he continued.
Parthasarathy also suggested that businesses should focus on designing the right user experiences, as it will “increase productivity and enable AI as a collaborator with humans (ensuring the user is empowered to benefit)”.
How do you expect to see the use of conversational AI evolve?
Vijay Parthasarathy, Head of AI/ML at Zoom
Parthasarathy advocated conversational AI to help workplaces “focus on meaningful and seamless collaboration”. “The little things add up to a lot of time and energy,” he said, “the time it takes to compose an email, find the notes about a call you missed, or catch up on unread chats. Organizations will adopt AI technology to empower their employees and improve their customers’ experiences.”
“The impact of AI will only expand the capabilities of digital tools to help us all work better and will help everyone to work smarter, not harder,” he concluded.
Noam Mor, Senior Manager, AI Product at Vonage
Mor cited how conversational AI could expand even more with generative AI tools like ChatGPT. While that’s likely the most immediate outcome, Mor believed “other areas of evolution will include integrations with other technologies like virtual and/or augmented reality”.
Mor suggested that improving language models will be a significant growth area as languages around the world continue to expand. “More focus will be put on natural language processing to help increase the ability to deliver personalized experiences to customers,” he said. “Conversational AI will continue to help human support staff offload simple tasks, but finding the balance between automation and the empathetic response of human agents to maintain the delivery of high-quality CX experiences will be crucial.”
Neal McMahon, Regional Sales Leader, UCaaS and CCaaS, Avaya
McMahon pointed out that AI-enabled tools are becoming increasingly vital to business communications beyond just the contact centre, whether it’s “streamlining tasks, providing personalised training, or delivering more inclusive and diverse customer service”.
“As the market continues to evolve, businesses that fail to capitalise on these benefits risk reputational damage and losing employees to brands that will fulfil these expectations,” McMahon commented. “So, it’s time to embrace the future with open arms and harness the power of conversational AI to stay competitive in a rapidly changing landscape.”
Joachim Jonkers, Conversational AI, Product Lead at Sinch
Jonkers argued that systems would continue to improve their understanding of natural language and context over the next few years. “They will become more adept at deciphering complex user queries, understanding nuances and providing accurate and relevant responses,” he added.
“Conversational AI will also evolve to support multimodal communication, combining voice, video, text and other forms of interaction,” Jonkers expanded. “Users will have the flexibility to switch between communication channels seamlessly, maintaining context and continuity across different modalities.”
Jonkers also suggested that future conversational AI systems could be designed to recognize and respond to human emotions, particularly through voice interactions: “Sentiment analysis techniques to gauge user emotions and adjust their tone, language, and responses accordingly. This empathetic approach can enhance user satisfaction, particularly in customer service and support scenarios.”
Lastly, Jonkers stressed conversational AI’s capability to advance in personalisation capabilities. “By leveraging data analytics, user preferences, and historical interactions, these systems will deliver highly tailored experiences,” Jonkers said.
from UC Today https://ift.tt/QmBpONY
0 Comments