Integrating DeepSeek, Continuously Building A+I (Automation + Intelligence) Operation Service Capabilities
2025-03-06
Since version 8.0, CloudChef Platform now supports a large model chatbot, continuously integrating large model capabilities to enhance user experience and reduce the daily workload of IT operation.Recently, CloudChef Platform has completed deep integration with the DeepSeek-R1 large model, supporting private model fine-tuning to ensure data does not leave the domain, meeting the high data security requirements of finance, state-owned enterprises, and government sectors.Below is how CloudChef Platform integrates with the DeepSeek-R1 large model:In the large model service of the CloudChef Platform, DeepSeek-R1 can be configured as a model supplier.After connecting with the DeepSeek large model, the CloudChef Platform can provide two new features: an intelligent assistant and conversational IT services.
1.Large Model Empowered Intelligent AssistantThe intelligent assistant upgrades the knowledge base from passive retrieval to active service, with main scenarios including:• When a user submits a ticket, the intelligent assistant automatically links cases and operational documents in the knowledge base, proactively providing possible solutions.• Users can directly converse with the intelligent assistant to obtain the help they need.The training data for the large model comes from the accumulated tickets, operation manuals, and daily operation data on the CloudChef Platform.2.Conversational IT ServicesThe CloudChef Platform has added a conversational interaction method on the original mobile App, Web, and mini programs. Users can interact with the AI Agent through natural language to directly apply for IT services or perform change operations, lowering the usage threshold of traditional service catalogs. For example, users can apply for resources through natural language commands, such as "Please build a test environment with 10-core CPU for the marketing department, pre-installed with MySQL5.7" or mixed Chinese-English commands "Add 16 vCPUs to the server of the big data group, upgrade memory to 128GB". After the Agent parses the command, it calls the SmartCMP API to execute the resource application or change operation. The entire process is based on a three-layer architecture of "intent recognition - entity extraction - strategy generation", with a semantic parsing accuracy rate of 98.2% for resource application scenarios.
Globally, ServiceNow has adopted AI large models as a key means of IT operation, striving to become the "control tower of AI business transformation". CloudChef Technology will also adopt more AI technologies to continuously transform customers' IT and enhance the operation and maintenance management system.
