| Studio | The Studio of Smart Vision provides users with a comprehensive and customized AI transaction processing environment. After users log in to Smart Vision, they can perform a series of AI tasks such as dataset management, model training, knowledge base management, AI application orchestration, workflow orchestration, prompt management, tool management, to meet the needs of diversified scenario applications. |
| Smart Vision Store | The Smart Vision Store includes Dataset Store, Model Store, Capability Store, Tool Store, and Application Store. It provides multi-domain, multi-scenario, and popular public resources/components. Users can add suitable resources and services from Store to the Studio according to their own needs and characteristics to improve the efficiency of AI application development. |
| Settings | Through the setting function of Smart Vision, users can perform cloud native environment management, model call monitoring, log monitoring, user management, role management. Customized settings are used to meet the computing power, permissions and other requirements of different scenarios, helping users to flexibly respond to various business challenges. |
| Datasets | Dataset refers to a set of related data samples, usually used in fields such as machine learning, data mining, statistical analysis, etc. In Smart Vision, dataset is the basis for training your own proprietary large model. You can add public datasets from Dataset Store, or build private datasets that you can use to train customized models. |
| Large Model | Large models refer to deep learning models with a very large number of parameters, which are usually used to process large-scale datasets and complex tasks. You can add public large models from Model Store or create private large models to meet your individual needs. |
| NLP | Natural Language Processing refers to allowing the computer to accept input in the form of natural language from users, and internally perform a series of operations such as processing and calculation through algorithms defined by humans to simulate human understanding of natural language and return the results expected by the user, realizing the use of computers Replace manual labor to process large-scale natural language information. |
| Computer Vision | Computer Vision is a field of research that helps computers use complex algorithms, either traditional or deep learning-based, to understand digital images and videos and extract useful information. |
| multimodal | In the computer field, every source or form of information can be called a modality. Multimodal models refer to artificial intelligence models that can process and understand multiple different types of input data (such as text, images, audio, etc.). These models can learn and integrate the features of different modalities (i.e., data types). Perform a complex series of tasks. |
| Enterprise Knowledge Base | Enterprise Knowledge Base is a platform for centralized storage and management of knowledge within an organization. It helps enterprises effectively organize, store, retrieve and share various types of information and knowledge resources. You can create a knowledge base to facilitate AI applications to conduct questions and answers based on knowledge base files. |
| RAG | Retrieval-augmented Generation is a model that combines retrieval and generation technologies. It generates answers or content by referencing information from external knowledge bases. It has strong interpretability and customization capabilities and is suitable for question and answer systems, document generation, and intelligent assistants, and many other natural language processing tasks. |
| Embedding | Embedding is a method of mapping high-dimensional data to low-dimensional space. It is usually used to convert discrete, non-continuous data into continuous vector representation to facilitate computer processing. |
| AI Capability | AI capabilities refer to the functions and capabilities of large models when performing various tasks. They are provided to developers through SDK to conveniently develop related applications. They are mainly used to process the input and output of human-computer interaction, including audio and video processing, Text processing, etc. |
| SDK | Software Development Kit is a software package that provides development tools, documentation, and sample code for a specific platform or programming language. |
| API | Application Programming Interface is an interface used for mutual communication and interaction between software systems. It allows different applications to perform operations such as data exchange, function extension and service invocation, thereby realizing more complex and powerful applications. |
| API Key | The application interface key is a code used to verify and authorize API requests. It is a security credential that users need to use when accessing the API for authentication. This mechanism ensures that users can operate corresponding resources after successful authentication. |
| AI Application | AI Application is an application based on artificial intelligence technology, which has a wide range of application fields and huge development potential. You can add public AI applications from the Application Store or create private AI applications to meet your scenario application needs. |
| Workflow | AI Agent Workflow is a method of embedding AI agents into workflows to automate and enhance tasks, covering the entire process from task definition to result delivery, and leveraging the perception, reasoning, decision-making and execution capabilities of AI agents to improve efficiency and effectiveness. |
| Prompt | Prompt refers to the input instructions or keyword phrases provided when interacting with artificial intelligence, and is used to specify the tasks that AI should perform and the output it generates. Prompt is a bridge of communication between users and AI. By carefully designing and optimizing prompt, AI tools can be used more effectively to complete various tasks. |
| Load Balancing | A service that distributes traffic on demand. It expands the throughput capacity of the application system by distributing the traffic to different back-end servers. It can also eliminate single points of failure in the system and improve the availability of the application system. |
| Developer Space | You can discover the amazing large model application space of Smart Vision. You can also use the large model, knowledge base, AI capabilities, persistence layer services and other components of Smart Vision to create a unique space, publish and run it, and Share it with other users of the platform. |
| Cloud Native Environment | Create and manage the running environment of large models, which is a standard cloud-native Kubernets cluster. Smart Vision provides a default cluster, and the services running on it will have time limits. You can also create your own cluster, and the services running on it will be maintained for a long time. |
| Tenant | In a cloud-native environment, a tenant is a logical concept that represents a specific data/resource allocation unit. The applications of each tenant are isolated from each other and do not interfere with each other. There can be multiple users under one tenant. |
