# Function Overview Viking AI Search utilises advanced Large Language Model (LLM) understanding and multi\-modal embedding technologies. By combining the precision of traditional keyword matching with the semantic flexibility of vector\-based matching, it meets the diverse search requirements of e\-commerce, commercial asset repositories platforms, content platforms, and other scenarios:

More than keyword matching

![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/24864dff8afc4c5c9eece310b2a06e9d =1576x) AI search combines the precise query ability of keyword matching with the conceptual query ability of semantic matching, enabling natural language queries in the search box. While maintaining the accuracy of keywords, it enhances search generalization and diversity, improves matching for long\-tail search terms, and reduces empty searches.

Cross\-modal search

![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/1f8013346f604ba6a5fc1a00ab809c0e =1602x) The native cross\-modal understanding and matching ability of large models frees search engines from relying on heavy image keyword tagging. Simply import multimodal data to quickly build indexes and integrate multimodal search capabilities, including: * **Text cross\-modal search:** match each data item's image and text information with the user's query term simultaneously. * **Image\-to\-Image Search:** find similar images and product content. # Function Advantages * **Easy configuration, test immediately after data import:** Simply import data and configure fields to start testing search right away. It requires no coding, no heavy data labeling, or detailed index control. Leveraging large models' generalization ability supports content search across various industries. * **Native cross\-modal capability:** Items containing image data can utilise visual features for search out\-of\-the\-box. Support for text and image\-based queries is native, requiring no prior model training. * **Flexible configuration, adjust while experiencing:** Provides no\-code configuration for recall and ranking, combined with the **Experience Workstation** for instant strategy testing. Publish the strategy with one click once the results meet your requirements. * **Standardized deployment process**: Supports one\-click configuration launch and offers stable API interfaces for online business system integration, enabling rapid deployment and strategy iteration. # Start experiencing and configuring AI Search After creating an application and importing or linking an item dataset (refer to [Create an Application and Link Item Datasets](/docs/85296/1544946)), you can conduct search tests and fine\-tune results. Adjust search strategy parameters through the visual interface, monitor changes in search results in real time, and quickly optimize search performance. Click the 【Experience】 on the left side of the application interface to enter the experience workstation: ![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/7aa252f69a8743c29f2b4643d815a134 =4838x) You can also click the 【Experience】 button directly on the application page to enter the experience workstation: ![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/feb194651e5848bfb57c9e10ceff71a2 =3180x) Enjoy the search experience and adjust strategy configurations on the 【Search】 tab in the experience workstation: ![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/3a13f1bf263649579bd2b7669f393eab =4816x)