Mobile Purchasing Agent Automating Third-Party Checkout via ML
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulties in efficiently completing purchases on third-party merchant websites due to the need to manually input information and navigate through complex web interfaces, which can be time-consuming and error-prone.
Innovation Solution
A mobile application and server system utilize machine learning models to automate the purchase process by analyzing web page URLs and document object models to identify purchasable items and complete purchases on behalf of users, without requiring prior relationships with the merchant, by leveraging user-provided payment and shipping information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users manually input information and navigate through complex web interfaces to complete purchases, then they can execute purchases on third-party merchant websites, but the process becomes time-consuming and error-prone
Solution Approach 1:
The purchasing agent pre-fills form fields with user profile information (name, address, payment details) before the user completes the purchase, so that when the user initiates a purchase, the information is already prepared and ready to be submitted, eliminating the need for manual input during the purchase process
Solution Approach 2:
The system automatically navigates through the merchant website's purchase interface and submits the purchase request without requiring the user to manually click through each step, allowing the purchasing agent to autonomously complete the transaction based on user authorization
2Ease of operation
If users manually complete form fields and make link selections on third-party merchant websites, then they can customize their purchase, but the process becomes complex and error-prone
Solution Approach 1:
The purchasing agent acts as an intermediary between the user and the merchant website, translating user intent into automated actions on the website interface. It handles the complexity of navigating forms, selecting options, and submitting requests, while presenting a simplified interface to the user
Solution Approach 2:
The system replaces manual mechanical interactions (typing, clicking, navigating) with automated computer-controlled actions. The purchasing agent uses automated browsing and form-filling capabilities to perform tasks that would otherwise require manual user input and navigation
3Extent of automation
If a purchasing agent application automates the purchase process using machine learning models, then form field completions and link selections are automated, but the system requires access to user identity, payment and shipping information
Solution Approach 1:
The system is divided into distinct functional modules: a mobile application for user interaction, a server for coordinating purchases, machine learning models for form completion, and database components for storing user information. This segmentation allows each component to specialize in specific tasks and reduces overall system complexity
Data Source
AI summary
A purchasing agent application operates on a mobile device to facilitate a user's purchase of a product featured on a web page of a third party merchant for a gift recipient. The user can select an option in a web browser to share the web page URL with the application. The application communicates with a server to streamline the purchase process based on previously entered user identity, payment and shipping information. The application and/or server use machine learning models to automate the completion of any one or more form fields and/or link selections that the user would otherwise be required complete on the third party merchant's web site to execute the purchase. None of the user, the application or the server need have any prior relationship with the third party merchant. The machine learning models can be trained, for example, using supervised learning, unsupervised learning and/or reinforcement learning techniques.