📄 中文摘要
Totogi通过与AWS生成式AI创新中心合作，并利用Amazon Bedrock的快速创新能力，实现了对变更请求处理流程的自动化。该自动化解决方案的核心在于集成大型语言模型（LLMs）来理解和执行复杂的业务逻辑。具体而言，系统首先接收来自客户或内部团队的变更请求，这些请求通常以非结构化文本形式呈现。接着，利用Amazon Bedrock上部署的LLMs对这些文本进行自然语言处理，提取关键实体、意图和约束条件。例如，对于一个要求修改资费计划的请求，LLM能够识别出涉及的资费名称、修改类型（如价格调整、功能增减）、生效日期以及受影响的客户群。提取出的结构化信息随后被传递给Totogi BSS Magic平台，该平台作为业务支撑系统（BSS）的核心，负责管理电信运营商的计费、客户关系和产品目录。BSS Magic内部预置了丰富的业务规则引擎和API接口，能够根据LLM解析出的指令自动执行相应的业务操作。这些操作可能包括在产品目录中更新资费详情、修改客户订阅状态、配置新的计费规则或生成相关的通知信息。整个过程通过API调用实现无缝衔接，极大地减少了人工干预的需求。此外，该自动化框架还包含了错误检测和冲突解决机制。当LLM解析出模棱两可的指令或可能导致业务冲突的请求时，系统会触发人工审核流程，并向相关人员发送警报，确保业务操作的准确性和合规性。通过这种方式，Totogi不仅显著提升了变更请求处理的效率，将原本需要数小时甚至数天的人工操作缩短至数分钟，而且降低了人为错误率，释放了资源用于更具战略意义的任务。此方案展示了生成式AI在业务流程自动化方面的强大潜力，特别是在处理非结构化数据和复杂业务逻辑的场景下。
📄 English Summary
How Totogi automated change request processing with Totogi BSS Magic and Amazon Bedrock
Totogi has achieved automated change request processing by collaborating with the AWS Generative AI Innovation Center and leveraging the rapid innovation capabilities of Amazon Bedrock. The core of this automation solution lies in integrating Large Language Models (LLMs) to comprehend and execute complex business logic. Specifically, the system first receives change requests from customers or internal teams, which are typically presented in unstructured text format. Subsequently, LLMs deployed on Amazon Bedrock are utilized for natural language processing of these texts, extracting key entities, intentions, and constraints. For instance, for a request to modify a tariff plan, the LLM can identify the involved tariff name, modification type (e.g., price adjustment, feature addition/removal), effective date, and affected customer segments. The extracted structured information is then passed to the Totogi BSS Magic platform, which, as the core of the Business Support System (BSS), manages telecom operators' billing, customer relationships, and product catalogs. BSS Magic incorporates rich business rule engines and API interfaces, enabling it to automatically execute corresponding business operations based on the instructions parsed by the LLM. These operations may include updating tariff details in the product catalog, modifying customer subscription statuses, configuring new billing rules, or generating relevant notification messages. The entire process is seamlessly connected through API calls, significantly reducing the need for manual intervention. Additionally, this automation framework includes error detection and conflict resolution mechanisms. When the LLM parses ambiguous instructions or requests that may lead to business conflicts, the system triggers a human review process and sends alerts to relevant personnel, ensuring the accuracy and compliance of business operations. In this way, Totogi has not only significantly improved the efficiency of change request processing, shortening what previously took hours or even days of manual operation to minutes, but also reduced human error rates and freed up resources for more strategic tasks. This solution demonstrates the powerful potential of generative AI in business process automation, particularly in scenarios involving unstructured data and complex business logic.