Sobot is an integrated customer contact center solution provider, offering global enterprises integrated solutions based on the “customer contact center” scenario, including the three-dimensional integration of public domain + private domain, marketing + service, and software + BPO. Currently, it has provided services to thousands of leading enterprises in global retail e-commerce, finance, consumer goods, corporate services, education and training, as well as government and public utilities.
Sobot and Amazon collaborate on artificial intelligence
In the face of the “hallucination problem” of generative AI in the field of intelligent customer service, as well as the needs of multi-channel integration and interaction in complex scenarios, building on its previous generation of products, Sobot has adopted generative AI services such as AWS Amazon Bedrock and officially launched the Sobot AI Agent for customers. By leveraging AWS Amazon Bedrock to call large language models (LLMs), it introduces Retrieval Augmented Generation (RAG) to reduce the “hallucination” of the models.
Amazon System Architecture Diagram
Moreover, the knowledge base is upgraded to a “knowledge center”, allowing enterprises to maintain a unified knowledge center only, eliminating the need to configure different knowledge bases for different robots or manual workstations. Generative AI can be applied to all contact entrances. Through the unified knowledge center, it can understand, extract and generate responses, enabling all businesses to be empowered by generative AI.
Additionally, based on the knowledge of a single language version, the translation capabilities of large models can be used to automatically generate multi-language knowledge in different versions, enabling enterprises to easily handle business in multiple countries. Through intelligent questioning, intelligent collection by large models, and intelligent transfer to human agents, enterprises can configure personalized workflows in the existing service processes. Through intelligent task processes, robots can automatically conduct inquiries and handle responses throughout the process, forming a business closed loop.
In the end, Sobot has created a new intelligent customer contact model of large models that integrates “omnichannel application”, “high-accuracy response”, “light operation and maintenance”, and “personalized guidance”, further improving the customer service and operation efficiency of enterprises.
Sobot is a long-term partner of AWS. Based on generative AI services such as AWS Amazon Bedrock, it provides Mico with a global customer service contact center solution, enhancing the personalized service experience of VIP customers through the reception scenarios tailored to individual needs and the ability of hierarchical reception.
Project Background
Mico is a global open social platform dedicated to creating high-quality and rich open social products for global users, bringing a pleasant social experience, connecting people with each other and sharing happiness. Mico’s customer groups are complex, coming from multiple countries, speaking multiple languages and through multiple channels. It aims to use online customer service robots to solve common and general problems and improve the efficiency of agents. Different reception experiences are provided for customers at different levels, and differentiated services are offered. Active care and private operation services are provided for high-level VIP customers.
Solutions
• The capabilities of online customer service robots and Sobot AI Agent endow customers from different channels and in different languages with a more friendly dialogue experience.
• The ability of personalized reception for thousands of customers in thousands of ways of the online customer service better meets Mico’s need to distinguish different customers for personalized reception.
• The capability of email work orders supports the unified processing of emails from multiple mailboxes on one platform, with automatic transfer and cross-departmental collaboration.
Business Value
The accurate answer rate is over 80%, the efficiency is increased by 50%, and the customer satisfaction rate is over 95%.
