Sentient Assist and Automate (SAA)
Sentient Assist and Automate (SAA)
批准号:
133515
负责人:
金额:
$8.92万
依托单位国家:
英国
项目类别:
Feasibility Studies
财政年份:
2018
资助国家:
英国
项目状态:
已结题
起止时间:
2018 至 --
中文摘要
在英国和全球范围内,几乎70%的客户联系都是语音。在这些语音互动中,84%的人因为长时间的等待而感到沮丧,81%的人与不友好的员工打交道,80%的人不得不向几家公司的代表重复你的问题(来源:埃森哲2013年全球消费者脉动调查全球和美国主要调查结果)。Sentient Assist and automated是一个自我学习的语音机器人平台,旨在通过减少高达95%的等待时间、消除重复问题、提高呼叫中心的效率来消除这种沮丧情绪。该平台旨在实现大多数客户语音交互的自动化。对于剩余的呼叫,我们的目的是将其委派给相关的座席并协助他们,从而提高客户的利益,从而提高客户的满意度。为了实现这一目标,我们使用深度学习和自然语言处理的最新进展,从现有的对话中学习,以构建一个语音机器人,它强调并能够实时为客户解决问题。
英文摘要
Almost 70% of all customer contact in the UK and globally is voice. During these voice interactions, much frustration is caused by being put on hold for a long time - 84%, dealing with employees who are unfriendly - 81%, and having to repeat your issue to several company reps - 80% (Source: Accenture 2013 Global Consumer Pulse Survey Global & U.S. Key Findings).Sentient Assist and Automate is a self-learning voice-bot platform that aims to remove this frustration by cutting the waiting time by up to 95%, removing the repetition issue, and improving the effectiveness of the call centres. The platform aims to automate the majority of customer voice interactions. For the remaining calls, the aim is to delegate them to the relevant agents and assist them, so they improve the benefits for the customer, leading to higher customer satisfaction.To achieve this we use the latest advances in deep learning and natural language processing to learn from the existing conversations, in order to build a voice bot that is emphatic and capable of solving issues for customers in real-time.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
海外基金