CMSWIRE - Key Persons
Job Titles:
- President and Co - Founder of Akorbi
Azam Mirza, president and co-founder of Akorbi, a multilingual digital transformation group, told CMSWire that generative AI has been around for a few years now in the contact center industry. "Initial use cases were around emotion deduction based on tone or voice, speed of conversation, and choice of words used," said Mirza. "Then we started seeing trends in bots which are learning while being used." Mirza said that generative AI, which uses natural language processing (NLP), not only learns while an agent is on the phone, but can also function as a chatbot that has learned from all the data it received while listening to the calls.
Damien Thioulouse, head of AI at Simplr, a US-based customer service outsourcing platform and service provider, told CMSWire that generative AI has pushed the boundaries of what was possible in terms of automation. "Bots will be able to successfully engage with customers well beyond the single intent inquiries and also within more intricate applications such as technical troubleshooting," said Thioulouse. "Large Language Models are excellent at summarizing, organizing, and prioritizing topics. This is having a profound impact on using the customer's voice as input."
Dom Nicastro Dom Nicastro is an award-winning journalist and radio personality based in Manchester, NH. He currently serves as editor-in-chief for Simpler...
Job Titles:
- Head of Data Science at Usetech
Ilya Smirnov, head of data science at Usetech, a tailored software development company, told CMSWire that we rarely think about the fact that people working in contact centers are faced with a large amount of information, for the processing of which huge resources are used. Smirnov suggested that modern machine learning technologies can help to significantly reduce the cost of providing services, as well as improve the efficiency of the contact centers.
The most primitive robotic systems, said Smirnov, are linear chatbots - the type most of us have been used to seeing for many years now as we navigate the web. "We meet them in messengers, social networks, mobile applications and websites," Smirnov said. He added that these chatbots are not trained and work according to a certain scenario - but they are useful. "With their help, you can order a pizza or a table in a restaurant, specify the cost of sending a parcel or get a ticket to the doctor. At the same time, the average request processing time will be reduced by about 3 times," said Smirnov.
Smirnov worked with a company that was engaged in insurance and wanted to create its own first-line chatbot using AI to improve the quality of service and unload an overburdened call center. By the final stage of the project, the AI-driven chatbot was able to completely close 30% of requests in automatic mode without operator participation, while another 35% of requests required only operator confirmation. "As a result, the average time to solve an application in which the operator's participation is required was reduced from several hours to 10-15 minutes." It's clear today that AI can reduce the need for agents to directly communicate with customers, allowing them to be available for the most complex or detailed tasks.
Jennifer Torres Jennifer Torres is an award-winning journalist based in Florida. She formerly served as staff reporter for Simpler Media Group/CMSWire covering...
Joe Bradley, chief scientist at LivePerson, a conversational AI platform provider, told CMSWire that generative AI and LLMs such as ChatGPT are not ready for enterprise use right out of the box, and that there are several issues that must be addressed when using generative AI. "LLMs can include bias, whether through toxicity, hurtful language or polarizing responses leading to unintended consequences." This has already been seen in the past with Amazon's employment pre-screening AI, and more recently with the new AI-driven Bing.
Additionally, generative AI models must be fine-tuned using data that is relevant to the task, brand or industry. "Using LLMs to generate content that is relevant to your business with little or no customization or fine-tuning will result in responses that are too generic or even irrelevant to the brand's products or the consumer needs," said Bradley.
If generative AI is going to be used with workflows that are not included in the contact center solution a brand is using, it will have to be integrated with a brand's current software and procedures. "If a company wanted to use an LLM to generate automated responses to customer complaints, the LLM would not inherently understand how to integrate with the company's complaint management system and follow the appropriate workflow," said Bradley.
Similarly, while generative AI is great at having conversations, it does not innately have the ability to follow up with specific actions. "Large language models, such as GPT-3 or BERT, are designed to analyze and generate humanlike language. However, they are not typically connected to inventory systems and do not have the capability to perform orders, order tracking or checking if something is available for customer service," Bradley explained.
Job Titles:
- Principal, Unified Commerce Strategy at Perficient
Justin Racine Justin Racine is Principal, Unified Commerce Strategy at Perficient, a global digital consulting firm serving enterprise clients throughout North America...
Liraz Margalit Liraz Margalit, PhD, is a digital psychologist, customer & user behavior specialist, and an international keynote speaker. She integrates cognitive...
Job Titles:
- Senior Editor for Simpler Media Group
Michelle Hawley Michelle Hawley is a senior editor for Simpler Media Group and a reporter for CMSWire, Reworked and VKTR. She reports...
Scott Clark is a seasoned journalist based in Columbus, Ohio, who has made a name for himself covering the ever-evolving landscape of customer experience, marketing and technology. He has over 20 years of experience covering Information Technology and 27 years as a web developer. His coverage ranges across customer experience, AI, social media marketing, voice of customer, diversity & inclusion and more. Scott is a strong advocate for customer experience and corporate responsibility, bringing together statistics, facts, and insights from leading thought leaders to provide informative and thought-provoking articles. Connect with Scott Clark:
Job Titles:
- EVP of Product at Zendesk