Image-Based LLM Promotion Generation for Store Item and Price Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The process of creating promotional content for small or midsize source locations is inefficient and resource-intensive, as it typically requires manual or third-party integration, which does not scale well without dedicated budgets or employees.

Innovation Solution

An online system uses a large language model to automatically generate promotional content by extracting information from images captured at a source location, identifying items, their prices, and promotions, and generating content based on a database of items without human intervention.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual or third-party integration methods are used to create promotional content, then the content can be created with existing tools, but the process is inefficient and resource-intensive requiring multiple manual steps and dedicated employees

Engineering Contradiction:

Improvepromotional content creation efficiencyVSAvoidprocess complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system enables automated self-service promotional content creation by capturing images at source locations, automatically extracting item information and pricing data, and generating promotional content without requiring manual intervention or dedicated content creation teams. This resolves the contradiction by making the system self-sufficient while improving productivity.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical processes (human evaluation of inventory, price comparison, item selection, and content creation) with automated computational systems including image recognition algorithms, data extraction models, and content generation systems. This substitution dramatically improves efficiency while reducing process complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Adaptability or versatility

If small or midsize sources without dedicated resources attempt to create promotional content manually, then they can produce content, but the process does not scale well and requires disproportionate resources

Engineering Contradiction:

Improvescalability for small and midsize sourcesVSAvoidcontent creation output

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system provides a universal automated solution that can be deployed by source locations of any size without requiring dedicated content creation resources. The same automated pipeline that captures images, extracts data, and generates promotional content works equally well for small, midsize, and large sources, enabling scalability across different organization sizes.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent introduces an automated content generation system as an intermediary between source location inventory data and promotional content output. This intermediary handles all the complex processing steps automatically, allowing small and midsize sources to achieve scalable content creation without needing to build complex internal processes.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If multiple manual steps are used including inventory evaluation, price comparison, item selection, and template placement, then comprehensive promotional content can be created, but the process is time-consuming and requires human intervention at each step

Engineering Contradiction:

Improvepromotional content accuracyVSAvoidcontent creation time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary automated actions including capturing images of items at source locations, pre-processing the images, and pre-extracting item information and pricing data before content generation. This preliminary automated processing ensures accuracy while dramatically reducing the time required compared to manual execution of each step.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent merges multiple separate manual steps (inventory evaluation, price comparison, item selection, template placement) into a single integrated automated pipeline. The system simultaneously processes images, extracts relevant information, compares prices, selects items, and generates promotional content in one continuous automated flow, maintaining reliability while eliminating time loss.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS20260065327A1Using a large language model to generate content based on images captured at a source location

Publication Date: 2026.03.05 MAPLEBEAR INC

AI summary

An online system receives an image captured at a source location, in which the image depicts one or more objects. The system generates a prompt including the image and a request to identify, from the objects, a set of items available at the source location based on a database of items available at the source location, and to extract, from the image, text describing a price or a promotion associated with each identified item. The system provides the prompt to a large language model to obtain an output, in which the model is fine-tuned based on the database of items. The system extracts, from the output, an identifier and the text associated with each item, retrieves item data for each item based on the identifier associated with the item, and generates promotional content for the source location based on the item data and the price or promotion associated with each item.