Help Center/ PanguLargeModels/ Service Overview/ Functions/ Agent Development
Updated on 2025-06-30 GMT+08:00
View PDF
Share

Agent Development

The Agent development platform provides a comprehensive tool set to help you efficiently develop, optimize, and deploy agents. Whether you are a beginner or an experienced developer, you can quickly develop and deploy agents using the prompt engineering, plug-in extension, flexible workflow design, and full-link debugging functions provided by the platform, accelerating the innovation and deployment of industry AI applications.

  • For zero-code developers who have no coding experience:
    • The platform provides functions such as prompt engineering and plug-in customization, enabling you to quickly build, fine-tune, and run your own large model applications without writing code. Through simple configuration, you can effortlessly create an agent application and conveniently explore AI applications.
    • The platform provides a knowledge import function that enables you to store and manage data and interact with AI applications. Local documents in various formats, such as .docx, .pptx, and .pdf, can be easily imported into knowledge bases, providing personalized data support for agent applications.
    • The platform also offers comprehensive observation and debugging tools to enable developers to thoroughly analyze each phase of the agent's execution process. Hierarchical information display helps developers optimize the performance and stability of AI applications and ensures smooth operation of the applications across different environments.
  • For low-code developers (who have certain coding experience):
    • Based on the preceding functions, the platform also provides flexible workflow design capabilities, allowing you to compile minimal code to develop sophisticated and highly stable agent applications. Developers can swiftly set up workflows and achieve more efficient application development by combining various components, such as foundational models, code, and intent recognition, through a drag-and-drop interface.
    • The platform also offers comprehensive observation and debugging tools to enable developers to thoroughly analyze each phase of the workflow's execution process. Hierarchical information display helps developers optimize the performance and stability of AI applications and ensures smooth operation of the applications across different environments.
Parent topic: Functions