Overview

Introduction

Smart Vision is an Enterprise-level Agent platform that integrates the four key elements required for agent implementation: compute, models, knowledge, and applications. It addresses the engineering challenges across the full lifecycle of building, training, using, and managing agents, forming a complete pipeline for agent application engineering and effectively resolving the challenges enterprises face when deploying agent applications.

Product Architecture

What Smart Vision can do?

1. Customized AI Applications: Smart Vision can provide customized AI Application functions, namely Applications. Enterprises can build customized AI applications based on their own needs and characteristics to meet different business needs.
2. Provide Developer Space: Smart Vision can provide a block for managing AI applications, that is, a developer space. This space can help enterprises better manage and maintain AI applications, including operations such as deployment, monitoring, debugging, and upgrades, and improve the reliability and security of AI applications.
3. Support Develop components: Smart Vision can provide intermediate processing capabilities, namely Develop components. These capabilities can help the machine better understand and process methods, improve the intelligence level of the machine, and thus better support scene applications.
4. Workflow orchestration: Smart Vision supports workflow orchestration based on Agent IDE, namely Workflows. You can freely arrange and adjust, and flexibly configure each task node in the process and its execution logic to meet your diverse needs.
5. Build Enterprise Knowledge Base: Smart Vision can provide a platform for centralized storage and management of knowledge within the organization, that is, Knowledge. The knowledge base can include various knowledge, experiences and materials within the enterprise, helping employees better acquire and utilize knowledge and improve work efficiency.
6. Enterprise Data Asset Governance: Smart Vision supports the construction of enterprise data assets through Database Connection and enterprise data, based on data tables, indicators&dimensions, terminology&dictionaries, and associated data cabins, to build semantic models, achieve automated data processing and intelligent analysis and prediction, and unleash the commercial value of data assets.
7. Dataset Management: Smart Vision can provide the function of Datasets. This function can help enterprises better manage and utilize data sets, and provide a basis for you to train your own proprietary large models.
8. Model Training: Smart Vision can provide model training capabilities to solve your need to generate a model that meets your personalized needs from a general large model, that is, Model Training. This capability can help enterprises better train their models and improve the accuracy and reliability of the models, thereby better supporting business decisions and intelligent applications.
9. Provide More Choices: Smart Vision provides Dataset Square, Model Square, External servers, MCP Servers and Application Square functions to allow you to obtain popular public resources in the industry on demand, providing you with a greater choice. You can obtain suitable resources and servers from these functions based on your own needs and characteristics to improve the efficiency of developing AI applications.

After obtaining the user permissions of Smart Vision, Loginopen in new window Smart Vision (standard or LDAP or SSO), you can start using the corresponding role (for details, see: Roles) function of Smart Vision.