Store Server for Location-Aware Personalized Item Promotion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generative AI-based sales promotions on cloud systems face challenges in providing personalized and optimized item recommendations due to lack of customer data storage and lack of store-specific optimization, leading to potential misofferings.
Innovation Solution
A store server system that includes a network interface, memory, and processor to identify customers via image analysis, acquire location information, and generate personalized promotional text using a machine learning model trained for store-specific item promotion, ensuring appropriate item recommendations based on customer attributes and location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a cloud computing system is used for generative AI-based promotion service, then the system can process promotion requests from multiple customers, but customer personal information cannot be stored on the cloud system when privacy restrictions are applied
Solution Approach 1:
The system segments the promotion service into two parts: a cloud-based generative AI processing component that handles promotion request processing, and a local store server component that stores customer personal information. This segmentation allows the system to maintain promotion service capability while storing customer data locally at the store, resolving the contradiction between cloud processing versatility and local data storage requirements.
2Productivity
If a cloud-based generative AI model is used for promotion service, then the system can provide promotion recommendations, but the model is not optimized for individual store characteristics leading to misofferings
Solution Approach 1:
The system implements local quality by training separate generative AI models at each store based on store-specific item data and customer behavior patterns. This allows each store to have an optimized promotion model that reflects its unique characteristics, improving recommendation accuracy while maintaining the overall promotion service capability through the cloud-based framework.
3Speed
If a general generative AI model is used for promotion service, then the system can generate promotion text quickly, but the generated text may not be appropriate for the specific store context or customer preferences
Solution Approach 1:
The system performs preliminary action by pre-training store-specific generative AI models using historical store data, item information, and customer preferences before actual promotion requests are processed. This pre-training enables the models to generate relevant and appropriate promotional text quickly when actual promotion requests are received, resolving the contradiction between generation speed and text relevance.
Data Source
AI summary
A store server for managing data of items sold in a store, includes a network interface connectable to a customer terminal in the store, a memory, and a processor configured to execute a program stored in the memory. The program causes the server to: upon receipt of an image from the terminal, identify a customer, and acquire customer information corresponding thereto, upon receipt of location information from the terminal, determine a location of the customer in the store, generate first text indicating the location and attributes corresponding to the customer information, input the first text to a machine learning model trained to generate item text indicating an item sold in the store and to be promoted, and generate second text for promoting a first item based on item text output from the model, and control the network interface to transmit the second text to the customer terminal.