ARTIFICIAL INTELLIGENCE CONTRIBUTION DETERMINATION SYSTEM

Abstract

This invention relates to a system (1) that enables transparent monitoring, reporting and auditing of the artificial intelligence contribution rate in call center and CRM (Customer Relationship Management) infrastructures.

Description

1 TARIFF ARTIFICIAL INTELLIGENCE CONTRIBUTION DETERMINATION SYSTEM Technical Area This invention enables artificial intelligence in call center and CRM (Customer Relationship Management) infrastructures. transparent monitoring, reporting and auditing of the intelligence contribution rate It is related to a system that provides. Previous Tech 10 Today, call centers offer solutions to the problems of customers who call them. Artificial intelligence is used in its creation, but to what extent is artificial intelligence utilized? No information regarding the use of this service is shared with the customer. This situation... This undermines customer trust. 15 Call for proposals for the invention with application number 2021 / 885228, which is included in the prior art. central solutions use artificial intelligence (AI) as a recommendation engine or summarizer. It uses AI, but the customer is not given transparent information about the use of AI. The contribution is not logged separately from the representative's contribution, in accordance with legal regulations (KVKK, 20 GDPR) appropriate transparency mechanisms are not in place and Ledger-based No immutable record keeping is performed. Therefore, the aforementioned deficiencies... A new system is needed to overcome this. Brief Description of the Invention 25 The purpose of this invention is to improve call center and CRM (Customer Relationship Management) systems. transparent monitoring and reporting of the artificial intelligence contribution rate in their infrastructures. and to create a system that ensures its monitoring. 2 Detailed Description of the Invention The "Artificial Intelligence Contribution Determination" process was carried out to achieve the purpose of this discovery. The "System" is shown in the attached figure; Figure 1. Schematic view of the system described in the invention. The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. given below: 1. System 2. Electronic device 3. Server The contribution rate of artificial intelligence to call center and CRM infrastructures should be transparently set at 15%. The system in question, which enables monitoring, reporting and auditing (1); - customers can contact call centers and call center employees can interact with them. at least one electronic system configured to provide services to customers device (2) and -20 requests from customer or call center employee via electronic device (2) receiving the request, forwarding the received request to artificial intelligence, processing the incoming request... labeling which category they belong to, while receiving data from artificial intelligence. Calculating the contribution percentage based on the contribution, and assigning the contribution percentage to the customer. a statement card should be prepared for submission, and the statement card should include information regarding the request. The contribution of artificial intelligence to the solution produced and the reasons for using artificial intelligence 25 providing information to the customer and customer representative by explaining the situation. It includes at least one server (3) configured to provide. The electronic device (2) in the system (1) which is the subject of the invention, allows data input. providing at least one keypad or touchscreen, 30 customers calling the call center To what extent does artificial intelligence contribute to the solutions produced for these problems? 3 having at least one screen that displays information regarding the reasons, It is a device such as a phone, tablet computer, or laptop computer. The device in question... electronic device (2), any remote communication device in the known state of the art to establish a connection with the server (3) using the protocol and this established connection It is configured to exchange data with the server (3) via 5 In the preferred application of the invention, the electronic device (2), server (3) and Internet to exchange data using a data path in this manner It is being structured. The server (3) in the system in question (1) is one of the 10 in the known state of the art. to communicate with an electronic device (2) using any communication protocol and data exchange with electronic device (2) through this established communication It is configured to perform. Server (3), electronic device (2) With the customer's permission, the customer profile and history of the customer who called the call center are accessed. CRM data such as campaign responses and transaction records, customer representative 15 Transcripts of conversations with [the organization / individual], agent notes, AI (Automated Resource Storage). Call center data in the form of suggestion records, processed by generative artificial intelligence AI content data generated in the form of text, summary, recommendation or decision statement It is configured to enable the collection of data. The server (3) is configured to enable the collection of data. The immutable ledger 20, which has masked IDs compliant with the Personal Data Protection Law (KVKK). (Immutable Ledger) to be kept on record with a timestamp. It is configured to provide the server (3), call text, CRM context, transaction. Data in the form of type is in the form of GPT or T5 derivative encoder-decoder. Proposal accuracy rate and agent acceptance rate are processed using transformer-based LLM. 25 through monitoring metrics such as customer satisfaction score (CSAT) Outputs such as "text suggested to the representative" or "customer response prediction" It is structured to ensure that each output is produced by humans and To determine the AI ​​contribution rate, we use the model contribution tracker and human... When the input is 40% and the model completion is 60%, the label is "%60 AI contribution". It is configured to enable its creation. Server (3), history call type, 30 Result, agent action, CRM status data on XGBoost + Shapley 4 By conducting a value analysis, “The rationale for the action proposed in this call” → “Why: Producing outputs such as, "The same solution was successful in 124 similar calls," and this... to record the data along with the chain of rules on which each prediction is based. It is configured. Server (3) predicts the artificial intelligence model with SHAP / LIME. To derive the relationship between data features, which words can artificial intelligence identify as 5 Determine the focus using the Attention Heatmap and Model Attribution. With its log, it records the time, model version, and parameter set for each AI call. By identifying the "responsibility ID" data, we can understand why and how the AI ​​decision was made. It is structured to transparently explain the data used to obtain it. Server (3) transmits the disclosed data to customer 10 via the electronic device (2) interface. representatives' CRM screens or customer application / web pages It is configured to be reflected as a disclosure / transparency card. The server (3) stated that the explanation / transparency card reads, “This proposal was prepared with 60% AI contribution. Reason: Your financial history + similar complaints have been analyzed. Model: v3.1 | Ensuring that it includes information such as "Accuracy: 92% | Presented with human verification." 15 is structured. Server (3), each piece of content generated by AI, decision to ensure that it is recorded in an immutable ledger It is structured. The server (3) allows for the retrospective processing of decisions made with artificial intelligence. It is structured to enable the server (3) to be examined as an AI decision. If customer satisfaction has decreased, this data should be written into the model training set. It is structured to enable the development of the model by providing the server. (3), transparently showing when and to what extent AI contributes reporting, call center representative or customer's feedback regarding the solution provided being able to clearly see the source, model versions and reasoning behind the decision It is structured to enable monitoring. 25 Industrial Application of the Invention The invention subject system (1) artificial intelligence in call center and CRM systems It offers a layer that makes its use transparent. Which 30% of AI-generated content... how it is implemented in the processes, the extent of its contribution, and how it benefits the customer. All transactions are recorded and reflected in the logbook. It is stored with a timestamp. On the customer interface or agent screen, it says, "This suggestion..." Prepared with 60% AI input. Reason: Financial background + similar complaints” etc. Transparency cards are displayed. Around these basic concepts, the invention is called "Artificial Intelligence Contribution Determination System (1)" It is possible to develop a wide variety of applications related to this, and the invention is presented here. It cannot be limited to the examples given; it is essentially as stated in the claims.

Claims

6 REQUESTS 1. Transparency on the percentage of artificial intelligence contribution to call center and CRM infrastructures. enabling monitoring, reporting and auditing in this manner; - enabling customers to contact call centers and call center 5 structured to enable its employees to serve customers at least one electronic device (2) and - via electronic device (2) customer or call center employee receiving the request, forwarding the received request to artificial intelligence, incoming the requests are processed and tagged according to their category, during which time 10 Calculating the contribution percentage based on the contribution received from artificial intelligence, contribution Preparing an information card to present the percentage to the customer, The explanation card details the contribution of artificial intelligence to the solution generated in relation to the request and By explaining the reason for using artificial intelligence, the customer and at least 15 configured to provide information to the customer representative. a system (1) characterized by containing a server (3).

2. At least one keypad or touchscreen that enables data entry, call artificial intelligence refers to solutions produced for the problems of customers who call the center. Information on the extent and reasons for the contribution of intelligence is 20 phones, tablets that have at least one screen that allows them to display images. electronic device (2) characterized by a computer or portable computer A system like the one in Request 1 (1).

3. Connect to the server (3) using any remote communication protocol 25 to establish and exchange data with the server (3) through this established connection characterized by an electronic device (2) configured to perform A system like the one in Request 1 or 2 (1). 7 4. Data exchange with the server (3) using a data bus in the form of the Internet. characterized by an electronic device (2) configured to perform a system like the one in Request 3 (1).

5. Communicate with an electronic device (2) using any communication protocol. to establish and to send data via electronic device (2) through this established communication Characterized by the server (3) configured to carry out the transaction. a system like any of the above-mentioned requests (1).

6. Permission of the customer calling the call centre via electronic device (2) 10 including customer profile, past campaign responses and transaction records. The CRM data shown in the image, the transcripts of the conversations with the customer representative ASR-based written documentation, representative notes, AI proposal logs. Call center data in this form, generated by generative artificial intelligence. 15 AI content data in the form of text, summary, recommendation or decision statement characterized by the server (3) configured to enable the collection a system like any of the above requests (1).

7. The collected data has masked IDs that comply with the GDPR. 20 timestamped on an immutable ledger server configured to keep records (3) a system like any of the above characterized claims (1).

8. Data such as call text, CRM context, and transaction type should be in GPT or T5 25 format. processed by a transformer-based LLM in the form of a derivative encoder-decoder recommendation accuracy rate, agent acceptance rate, customer satisfaction score "Recommended to the Representative" based on monitoring metrics in the form of (CSAT). generating outputs in the form of "text" or "customer response prediction" The above 30 is characterized by the server (3) configured to provide a system like any of the requests (1). 8 9. Model contribution to determine the human and AI contribution ratio of each output. Using trackers and with 40% human input, model completion is 60%. to ensure that the label “%60 AI contribution” is created when it occurs 5 of the above requests characterized by the configured server (3). a system like any other (1).

10. On historical call type, outcome, agent action, and CRM status data. By performing an XGBoost + Shapley value analysis, it was determined that "This proposal in the call..." "Justification for the action" → "Why: Same solution successful in 124 similar calls" It provided outputs such as "and based every prediction on this data." Server configured to record together with the rule chain (3) as in any of the above claims characterized by system (1).

11. The relationship between AI model prediction using SHAP / LIME and data characteristics. to extract which words the AI ​​focuses on Attention Identify using a Heatmap and Model Citation Log for each Time, model version, parameter set, "responsibility" for AI call. By identifying the "(responsibility) ID" data, we can determine why, how, and which 20% of the AI ​​decision was made. Server configured to transparently explain that it was received with data (3) as in any of the above claims characterized by system (1).

12. The disclosed data is sent to customer 25 via the electronic device (2) interface. representatives' CRM screens or customer application / web pages configured to be reflected as a disclosure / transparency card in any of the above requests characterized by the server (3) such a system (1). 9 13. The disclosure / transparency card states, "This proposal was prepared with 60% AI input." Reason: Your financial history + similar complaints have been analyzed. Model: v3.1 It should contain information such as: "Accuracy: 92% | Presented with human verification." to provide the above characterized by the configured server (3) a system like any of the requests (1). 5 14. All content generated by AI is recorded in the immutable ledger (Immortality Register) of the decision. Server configured to enable recording (Ledger) with (3) a system like any of the above characterized claims (1). 10 15. Retrospective review of decisions made using artificial intelligence. the above characterized by the server (3) configured to provide a system like any of the requests (1).

16. If an AI decision has reduced customer satisfaction, this data should be added to the model training set. by enabling the writing of the code, thus facilitating the development of the model. from the above requests characterized by the configured server (3) a system like any other (1).

17. Transparently detailing when and to what extent AI has contributed. reporting, call center representative or customer to the solution produced being able to clearly see the source related to, model versions and decision the server (3) configured to enable the monitoring of the reasons A system like any of the above-mentioned described claims 25 (1).