Smart Negotiation Autonomous AI

Abstract

Traditional negotiation systems operate based on fixed rules, making it difficult to respond flexibly by effectively utilizing real-time market fluctuations and past transaction history. As a result, transactions often do not proceed smoothly, are limited to price negotiations, and make it difficult to offer value-added proposals (such as free shipping or extended warranties), thus hindering effective improvement of transaction success rates and user experience. [Solution] This invention provides a system that automates price negotiation and value-added proposals by combining generative AI and ensemble learning technology, utilizing real-time market data and transaction history. This enables flexible negotiation strategies that were difficult to achieve with conventional systems, resulting in an improved negotiation success rate. Furthermore, it can be easily integrated into existing systems via API, and implementation is simple.

Description

Technical Field

[0001] The present invention relates to an autonomous negotiation system that combines ensemble learning technology and generative artificial intelligence (AI). In particular, it relates to a technology in which AI autonomously executes price negotiation and value-added proposal in consumer-to-consumer (C2C), business-to-consumer (B2C), and business-to-business (B2B) transactions, improving the success rate and efficiency of transactions. Furthermore, the present invention relates to applying ensemble learning to enhance prediction accuracy and the reliability of negotiation strategies using multiple models, thereby improving the accuracy and flexibility of price negotiation.

[0002] The technical field of the present invention particularly includes technologies for dynamically optimizing price setting and negotiation processes by leveraging AI in e-commerce platforms and free market systems. By applying ensemble learning, it aims to achieve a highly flexible and real-time negotiation strategy that is difficult to realize with a single model, maximizing transaction efficiency. Furthermore, it belongs to the technical field of reducing the risk of halcyonization and leading to more stable transaction results through dynamic strategy adjustment using past transaction data and market data.

[0003] Specifically, a system that combines generative AI and ensemble learning analyzes market fluctuations and competitive trends occurring during transactions in real time, and multiple models cooperate to generate an optimal negotiation strategy. In this technical field, the application of ensemble learning not only improves the accuracy of negotiation strategies but also contributes to improving the reliability of the entire system. Also, having a structure that can be easily integrated with external systems via APIs and the ease of introduction in transaction platforms are also part of this technical field.

Background Art

[0004] Traditional negotiation systems were primarily designed as rule-based "AI that waits for instructions," and typically conducted price negotiations and proposals based on predefined conditions. These types of systems relied on fixed algorithms, making it difficult to flexibly respond to real-time fluctuating market conditions and customer needs. Furthermore, the success or failure of a deal depended heavily on predetermined price ranges and negotiation scopes, resulting in a narrow range of negotiation strategies and a low success rate.

[0005] Conventional technologies often focused solely on price reduction as a means of price negotiation, making it difficult to flexibly incorporate value-added offers (e.g., free shipping, extended warranty, special services). Furthermore, even negotiation systems using market data and transaction history had limited capabilities to dynamically optimize strategies using that data, often resulting in transactions proceeding without being appropriately adjusted based on past experience or current market conditions. As a result, transaction efficiency and satisfaction were often low for both sellers and buyers.

[0006] Furthermore, conventional AI negotiation systems typically relied on a single model for decision-making during negotiations, which presented a challenge: biases and errors stemming from the data and algorithms the model learned directly impacted the transaction outcome. In particular, there was a risk of "hallucination," where the AI ​​would provide incorrect information or make wrong judgments, hindering the establishment of a reliable transaction process.

[0007] To address these problems, technologies utilizing generative AI emerged, and autonomous negotiation systems based on transaction and market data began to be developed. However, generative AI alone still remained dependent on a single model, and it was not possible to completely eliminate bias and risk in negotiation results. Furthermore, the difficulty of reflecting complex market trends and transaction history in real time remained.

[0008] Ensemble learning techniques have gained attention as an effective solution to these challenges. Ensemble learning is a technique that utilizes multiple different models simultaneously and integrates their prediction results to provide accuracy and reliability that cannot be achieved with a single model. As a result, multiple models make predictions based on different data sources and patterns, and by integrating them, trading results become more accurate and stable. Furthermore, ensemble learning makes it possible to diversify the risk of hallucination and improve the reliability of negotiation systems.

[0009] Conventional technologies suggest that by using ensemble learning, which combines multiple models while maximizing the advantages of generative AI, it is possible to improve the accuracy and flexibility of price negotiations and develop dynamic negotiation strategies based on market data and transaction history. This is expected to result in higher transaction completion rates and user satisfaction than conventional systems that rely on a single model. [Prior art documents] [Patent Documents]

[0010] [Patent Document 1] Public Relations for Patent No. 6435064

[0011] [Patent Document 2] Public Relations for Patent No. 6728932 [Overview of the project] [Problems that the invention aims to solve]

[0012] Traditional negotiation systems primarily relied on "instruction-waiting AI" based on fixed rules, making it difficult for these systems to flexibly respond to real-time fluctuating market conditions and customer needs. In particular, they were often limited to price negotiations only, failing to incorporate value-added offers (e.g., free shipping, extended warranty periods, special services), frequently resulting in lower transaction success rates and customer satisfaction. Furthermore, traditional generative AI systems relied on a single model, leading to problems such as model bias and erroneous judgments due to hallucination negatively impacting transaction outcomes.

[0013] Furthermore, conventional technologies had fixed negotiation strategies and struggled to dynamically respond to past transaction data and market fluctuations, resulting in a limited success rate for transactions. This made it difficult to improve the user experience and maintain market competitiveness. In particular, there was a need for a flexible system that could respond quickly and effectively to sudden market changes and competitor product developments during transactions.

[0014] Furthermore, systems based on a single AI model often resulted in overly rigid pricing and negotiation strategies, making it impossible to dynamically negotiate while considering multiple factors. As a result, transactions were less likely to be completed, and there was a higher possibility of unfavorable transaction conditions for both sellers and buyers. [Means for solving the problem]

[0015] This invention solves these problems by providing an autonomous negotiation system that combines generative AI and ensemble learning. By using ensemble learning, prediction results obtained from multiple different models are integrated to achieve more accurate and flexible price negotiation and value-added proposals. Furthermore, because it can analyze market data and past transaction history in real time and dynamically adjust strategies, it can respond quickly to fluctuating market conditions without relying on fixed strategies like conventional systems.

[0016] Furthermore, this invention makes it possible to incorporate value-added offers into negotiations. By offering value-added services such as free shipping, extended warranties, and additional services, rather than relying solely on price reductions, it is possible to provide attractive transaction terms to buyers and improve the success rate of transactions. In addition, through ensemble learning, the AI ​​learns from past transaction data and customer purchase history, enabling it to develop more accurate strategies for future transactions.

[0017] The system of this invention can be easily integrated into e-commerce platforms and other systems via APIs. This eliminates the need to develop complex AI models, enabling rapid deployment and operation. Furthermore, the system can autonomously learn and optimize its strategy with each negotiation, improving the efficiency and success rate of transactions. [Effects of the Invention]

[0018] The autonomous negotiation system combining generative AI and ensemble learning according to the present invention enables highly accurate and flexible negotiation strategies that were not possible with conventional "instruction-waiting AI" or single AI models. By integrating prediction results from multiple models through ensemble learning, this system improves the accuracy of price negotiation and value-added proposals compared to conventional technologies, and significantly increases the success rate of transactions.

[0019] Furthermore, because the AI ​​can autonomously adjust negotiation strategies based on real-time market data and past transaction history, it is particularly noteworthy that it can respond quickly to rapid market fluctuations and price changes of competing products compared to conventional fixed negotiation methods. This increases the chances of completing transactions and allows for more attractive and optimal transaction terms for both sellers and buyers.

[0020] The present invention also enables incorporating value-added proposals (e.g., free shipping, extended warranty period, provision of special services, etc.) into price negotiations. As a result, the transaction conditions do not solely depend on a simple price reduction, and more flexible and attractive proposals can be made to the purchaser, expecting an improvement in the user experience. Thereby, user satisfaction is improved, and an effect of promoting customer repeat purchases can be expected.

[0021] By using ensemble learning, multiple models perform analyses based on different data respectively, and finally an integrated prediction result is presented. Therefore, accuracy and reliability that are difficult to achieve with a single model can be ensured. Thereby, the risk of hallucination is reduced, and the reliability and stability of the system are improved. Also, as the AI continues to learn transaction history and market trends, its accuracy increases with each negotiation, so the success rate of transactions continues to improve in the long term.

[0022] The present invention can be easily integrated into existing e-commerce platforms and other systems through APIs and can be introduced and operated in a short period. It does not require the development of complex AI models, enabling companies and individuals to easily utilize a highly accurate negotiation system, expecting a reduction in introduction costs and an improvement in efficiency. Therefore, it is highly convenient for sellers and purchasers, and an effect of maximizing transaction efficiency can be obtained.

[0023] Also, as the AI performs autonomous learning, the system is constantly evolving and optimizing negotiation strategies based on transaction data and customer behavior data. As a result, the overall performance of the system is improved, and in long-term operation, it is possible to simultaneously achieve maximization of transaction efficiency and improvement of user satisfaction.

Brief Description of the Drawings

[0024] [Figure 1] It is an overall system configuration diagram of the smart negotiation autonomous AI of the present invention.

[0025] [Figure 2] It is a flowchart in the smart negotiation autonomous AI of the present invention. [Modes for carrying out the invention]

[0026] The autonomous negotiation system utilizing generative AI and ensemble learning according to the present invention autonomously negotiates prices and proposes added value to both sellers and buyers on an e-commerce platform. This system consists of multiple elements, including a generative AI module, an ensemble learning module, a data collection module, an API connection module, and a user interface module.

[0027] The generative AI module plays a key role in price negotiation and value-added proposals in transactions. This module analyzes past transaction history, customer purchasing behavior, and market fluctuation data in real time to formulate optimal negotiation strategies. Furthermore, the generative AI automatically incorporates value-added proposals (such as free shipping, extended warranty periods, and special services), offering not only better price negotiations but also more attractive transaction terms.

[0028] The ensemble learning module improves prediction accuracy and reliability by combining multiple AI models. This module integrates multiple prediction results generated using different algorithms and data sources to optimize the final negotiation strategy. This enables dynamic and flexible negotiation, which was difficult with traditional single-model-dependent systems.

[0029] The data collection module is responsible for collecting market data and transaction history in real time and providing it to the generative AI and ensemble learning modules. This data collection module works in conjunction with external market data sources to constantly obtain the latest price trends and information on competing products, enabling the entire system to respond quickly to market changes.

[0030] The API connection module is an interface that facilitates integration with external systems and platforms. This system can be easily integrated into existing e-commerce platforms and third-party applications via API, allowing sellers and developers to implement the system quickly.

[0031] The user interface module provides an intuitive interface for sellers and buyers to operate the system. Through this interface, sellers can adjust prices and value-added offers, and buyers can review and respond to presented negotiation terms. The interface visually displays real-time market data and negotiation progress to help users make informed decisions.

[0032] The system adjusts its negotiation strategy in real time based on market fluctuations and buyer responses that occur during negotiations. For example, if the price offered by the buyer does not reach the seller's desired price, the generating AI automatically proposes value-added offers (e.g., free shipping or extended warranty) to increase the likelihood of a successful transaction. It also has a function to adjust the content of the proposals based on the buyer's response, allowing for flexible responses tailored to the user's needs.

[0033] Furthermore, the system records transaction data and negotiation results, and learns autonomously. This allows the system to learn from past transactions and evolve to propose more effective strategies in future negotiations. In particular, by analyzing conditions that resulted in high transaction success rates and proposals that resonated well with buyers, and incorporating these insights into subsequent negotiations, the overall performance of the system continuously improves.

[0034] This system is highly convenient for both sellers and buyers, and serves as a powerful tool to maximize the success rate of transactions. Sellers can save time and effort through AI-driven automated negotiations while offering favorable transaction terms. Meanwhile, buyers can negotiate more flexible and attractive terms, making the entire system a win-win situation for both parties. [Examples]

[0035] This embodiment specifically describes how an autonomous negotiation system combining generative AI and ensemble learning operates. The following scenario involves price negotiation between a seller and a buyer on an e-commerce platform.

[0036] First, sellers register their products on the e-commerce platform and set their desired selling price, minimum acceptable price, and negotiable value-added offers (e.g., free shipping, extended warranty, special services). This information is stored in the system's database and analyzed by generative AI and ensemble learning modules. Based on the seller's settings, the system is ready to present the optimal negotiation strategy in real time when negotiations begin.

[0037] When a buyer expresses interest in a product and considers purchasing it, they initiate price negotiations with the seller through the system. Once the buyer proposes a desired price, the generating AI develops an optimal negotiation strategy based on past transaction data, market data, and the seller's minimum acceptable price. In this process, the generating AI can not only negotiate on price but also offer added value, such as free shipping or extended warranty periods.

[0038] For example, if the price offered by the buyer falls below the seller's minimum price, the AI ​​will try to close the deal by offering added value, such as free shipping or an extended warranty, to ensure the seller is not disadvantaged by the buyer's offer. In this way, instead of simply relying on price reductions, the AI ​​can significantly improve the deal completion rate by offering flexible proposals that combine added value.

[0039] Furthermore, this system can adjust its proposals in real time based on buyer reactions and market data during negotiations. For example, if a buyer reacts negatively to a presented proposal, the system immediately generates a new proposal offering more favorable terms. In this process, an ensemble learning module integrates prediction results from multiple AI models to determine the optimal terms for the buyer, enabling flexible and highly accurate negotiations.

[0040] Another feature of this system is that the results of negotiations are recorded and used for future transactions. The system learns patterns from successful and unsuccessful negotiations and develops more accurate negotiation strategies for subsequent transactions. For example, if free shipping is effective in closing a deal for a particular product category, the system can use that data to proactively propose free shipping in the next negotiation.

[0041] Furthermore, the system is designed to maintain an optimal balance at all times, ensuring profits for sellers while offering attractive terms to buyers. For example, it automatically selects price reductions that do not significantly reduce profit margins and value-added offers that do not incur additional costs, thereby increasing the likelihood of a successful transaction.

[0042] Furthermore, the system can be easily integrated into existing e-commerce platforms via APIs. By implementing this system, sellers will no longer need to manually adjust prices or offer value-added products, as AI will automatically handle negotiations. Buyers will also be able to easily review the offered terms and track the progress of negotiations in real time, enabling smooth and stress-free transactions.

[0043] In this embodiment, improvements in the accuracy of price negotiations and the transaction completion rate have been demonstrated. Furthermore, by incorporating value-added proposals, it is possible to offer attractive terms to buyers without relying on price reductions, thus protecting sellers' profits while increasing buyer satisfaction.

[0044] This system is not limited to specific industries or product categories and can be applied to a wide range of e-commerce platforms. For example, it is expected to be highly effective in the trading of various products such as clothing, home appliances, furniture, and digital content. [Industrial applicability]

[0045] The autonomous negotiation system combining generative AI and ensemble learning according to the present invention is applicable to a wide range of industrial fields. In particular, it offers significant advantages in automating price negotiation and value-added proposals on e-commerce (EC) platforms. This system can improve transaction success rates and customer satisfaction, making it an extremely useful tool for EC site operators and sellers.

[0046] This system can streamline the price negotiation process in consumer-to-consumer (C2C), business-to-consumer (B2C), and business-to-business (B2B) transactions. For example, on a C2C platform, AI automates negotiations between sellers and buyers, facilitating the completion of transactions. In B2C settings, it can dynamically set prices for products and services offered by companies and propose appropriate added value to customers, thereby increasing customer purchasing intent. In B2B, even in large-scale transactions and contracts, automating price adjustments and contract term negotiations can significantly improve the efficiency of business negotiations.

[0047] Furthermore, this system can be applied not only to e-commerce platforms but also to other industrial sectors. For example, in supply chain management, it can automate price negotiations and contract term adjustments with suppliers, supporting efficient inventory management. In the logistics industry, AI can adjust delivery costs and delivery dates, enabling real-time negotiations with trading partners.

[0048] The system of this invention is also effective in the insurance and financial industries. For example, AI can adjust and customize insurance contract terms, proposing appropriate insurance plans to customers. In the financial industry, AI can support negotiations on loan terms and investment product prices, providing customers with optimal proposals. This streamlines complex negotiation processes and improves the customer experience.

[0049] This system is also highly effective in the service industry. For example, in the hotel and airline industries, it can dynamically adjust accommodation and airfare prices in real time, allowing AI to offer optimal prices and value-added suggestions tailored to customer needs. This makes it possible to provide attractive conditions for customers while maximizing revenue for businesses.

[0050] This invention can be easily integrated into existing systems via APIs, does not require the development or implementation of complex AI models, and can be implemented in a relatively short period of time. Therefore, businesses of all sizes, from startups to large corporations, can adopt this system to streamline negotiation processes and improve transaction success rates.

[0051] Furthermore, the system of this invention autonomously learns based on data, and its accuracy improves with each negotiation, so its effects are sustained even in long-term operation. As a result, companies can improve the efficiency and profitability of transactions while reducing time and costs, contributing to strengthening their competitiveness.

[0052] Furthermore, this system also brings improved convenience to consumers. Through AI-driven flexible negotiation and value-added proposals, consumers can conduct transactions under conditions that suit their needs, resulting in a more satisfying purchasing experience. This encourages repeat business and leads to long-term customer acquisition for businesses.

Claims

1. A negotiation system that autonomously performs price negotiations and value-added proposals on an e-commerce platform using generative AI and ensemble learning, characterized in that it includes a generative AI module that collects market data and past transaction history and dynamically adjusts price negotiations and value-added proposals based on them.

2. The negotiation system according to claim 1, characterized in that the generating AI module uses an ensemble learning algorithm and integrates prediction results obtained from multiple different AI models to optimize price negotiation and value-added proposals.

3. A negotiation system according to claim 1 or 2, characterized in that the value-added offer includes free shipping, an extension of the warranty period, and the provision of special services.

4. A negotiation system according to any one of claims 1 to 3, characterized in that it dynamically adjusts the price negotiation strategy based on real-time market data.

5. A negotiation system according to any one of claims 1 to 4, characterized in that the generating AI has an autonomous learning function to analyze past transaction history and develop a more accurate strategy in the next negotiation.

6. A negotiation system according to any one of claims 1 to 5, characterized in that it can connect to an external e-commerce platform or system via an API and is capable of acquiring and processing negotiation data in real time.

7. A negotiation system according to any one of claims 1 to 6, characterized in that the ensemble learning algorithm improves the success rate of negotiations by combining multiple AI models using bagging.

8. A negotiation system according to any one of claims 1 to 7, characterized in that the ensemble learning algorithm uses boosting and optimizes the next negotiation strategy when an error occurs during negotiation.

9. A negotiation system according to any one of claims 1 to 8, characterized in that the generating AI calculates an optimal price in accordance with the price offered by the buyer and presents negotiation materials including value-added proposals.

10. A negotiation system according to any one of claims 1 to 9, characterized in that, when the price offered by the buyer during negotiations falls below the seller's minimum acceptable price, the system facilitates the completion of the transaction by adding value-added offers such as free shipping or an extended warranty period.

11. A negotiation system according to any one of claims 1 to 10, characterized in that it includes a user interface in which a generating AI visually displays the progress of negotiations and the terms of the transaction.

12. The negotiation system according to claim 11, characterized in that the user interface displays the progress of the negotiation strategy and the content of the proposals in real time to both the seller and the buyer, and the seller can easily manipulate price adjustments and value-added proposals.

13. A negotiation system according to any one of claims 1 to 12, characterized in that the generating AI makes individually optimized price proposals and value-added proposals based on the buyer's past purchase history.

14. A negotiation system according to any one of claims 1 to 13, characterized in that ensemble learning uses an optimization algorithm for making offers attractive to buyers while maximizing the seller's profits.

15. A negotiation system according to any one of claims 1 to 14, characterized in that, if a negotiation is not concluded, the generating AI analyzes the negotiation result and incorporates it into subsequent transactions.

16. A negotiation system according to any one of claims 1 to 15, characterized in that the generating AI presents the buyer with the optimal combination from among multiple value-added proposals in order to increase the likelihood of a transaction being concluded.

17. A negotiation system according to any one of claims 1 to 16, characterized in that the generating AI adjusts the strategy during negotiations based on the buyer's response to improve the success rate of the transaction.

18. A negotiation system according to any one of claims 1 to 17, characterized in that a generating AI evaluates the negotiation results using multiple transaction data and forms a feedback loop that reflects the results in future strategies.

19. A negotiation system according to any one of claims 1 to 18, characterized in that ensemble learning integrates data from multiple AI models based on past negotiation data and market trends to provide a strategy for maximizing the success rate of transactions.

Images