AN ORCHESTRATION SYSTEM

Abstract

This invention relates to an orchestration system (1) for reflecting context extracted locally on a customer-side end device to multiple customer interaction channels and services on a mobile network operator infrastructure.

Description

1 TARIFF AN ORCHESTRATION SYSTEM Technical Area This invention provides an endpoint on the customer side, over a mobile network operator's infrastructure. context extracted locally on the device can be applied to multiple customer interaction channels and It involves an orchestration system for reflecting this in the services. Previous Tech 10 In the current technology, interactions are carried out according to the customer's immediate context. The raw data used mostly includes location, movement, device status, and similar information. They are processed in centralized systems; this situation is problematic in terms of personal data protection. The risk involves additional data transfer and delays. Furthermore, the existing multi-channel 15 In customer interaction systems, the customer's immediate context influences channel selection decisions. SMS, email, push, IVR / feedback with low and predictable delay Interactions initiated across different channels, such as outgoing and incoming calls, share common and reciprocal characteristics. It is not possible to manage them in a way that can be achieved together. As a result... actions that occur after the context has lost its validity, overlapping or 20 Duplicate customer interactions can occur. Therefore, raw customer data... the ability to use the immediate context without removing it from the end device, this context Transfer of latency budget to the operator side and multi-channel ensuring that their interactions are carried out in a coordinated and, if necessary, reversible manner. A new system is needed to provide this. 25 United States patent number US12219027, which falls under the prior art. document 5G network slicing + URLLC + Multi-Access Edge Computing (MEC) by integrating orchestration and edge computing resources into low-latency services. A mechanism for allocation is being discussed. Here, the system involves network and computation 30 2 It focuses on source orchestration; context generated by native AI at the edge device. The concept of a token is absent; there is no multi-channel customer interaction selection logic. Brief Description of the Invention The purpose of this invention is to create a customer-side interface on a mobile network operator's infrastructure. context extracted locally at the end device can be applied to multiple customer interaction channels and The goal is to implement an orchestration system to reflect this in the services. Another aim of this invention is to create an endpoint device (mobile device, smart 10) on the customer side. IoT devices (connected in-vehicle interaction surfaces, etc.) utilize locally running artificial intelligence. A structured context token generated by an inference engine, suitable for 5G low-level applications. delayed communication infrastructure (network slicing, URLLC, or similar mechanisms) that it is transmitted via an orchestration engine on the operator side; orchestration the engine's context marker, interaction history, and channel-suitability scores are 15 by combining multiple customer interaction channels (SMS, email, push, IVR-feedback- (search, outgoing call, etc.) selects the appropriate one or ones; and the selected channel a common set of compensation definitions generated at the time of execution of triggers By being added to the queue, the customer can respond or the context can change, allowing the customer to re-enter the chain. The goal is to create a system from which withdrawal is possible. 20 Detailed Description of the Invention An "Orchestration System" was created to achieve the purpose of this invention. as shown in the figure; 25 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. The following are given below: 30 3 1. System 2. Electronic device 3. Artificial intelligence inference engine 4. Token transmission module 5. Multi-channel orchestration module 5 6. Undoable action module 7. Server On a mobile network operator's infrastructure, locally on a customer-side end device. The context extracted as such encompasses multiple customer interaction channels and services. The subject of the invention is the orchestration system used for its reflection (1); - used / carried by the customer and location, movement status, device status, Final application interactions, numerous signals via sensor outputs. at least one electronic device that enables its collection (2), -operated locally on the customer's electronic device (2) and 15 on the device (2) Signals include intent category, interest signal strength, availability status, and privacy tag. an edge AI that generates a context marker including validity and lifetime fields inference engine (3), -the context marker in question, raw customer signals within the device (2) limits by holding, 5G low latency communications infrastructure (network slicing, URLLC or 20 a deterministic delay budget allocated through (similar mechanisms) a token transmission module (4) that transmits to the operator side inside -the context marker in question, customer interaction history and each existing customer By combining it with the channel-suitability score for the interaction channel, at least for selection. A multi-channel 25 that selects one or more of two configurable channels. orchestration engine (5), - Each interaction message triggered from the selected channels is generated at the time of execution. Writing to a sequential queue, along with the definition of compensating action. when the customer responds through another channel or the context token is valid When its lifespan ends, the tail in question is reduced to 30 in a last-in, first-out order. a reversible action module (6) that carries out the compensatory steps by wrapping and 4 -electronic device (2), artificial intelligence inference engine (3), token transmission module (4), The single multi-channel orchestration engine (5) and the retractable action module (6) joint management of a customer interaction event and decision-making at each stage It includes at least one server (7) that enables writing to a persistent audit log. The electronic device (2) included in the system (1) which is the subject of the invention is a smartphone, smartwatch, An IoT device is any customer end device in the form of a connected in-vehicle interface. Electronic device (2), a tracker, device sensor or caller working on it. Current location from signals, whether it is in motion or not, device charge / battery Data such as status, active application category, and last interaction times are included in 10 to allow for collection and storage within the device's boundaries It is being structured. The artificial intelligence inference engine (3) included in the system (1) of which the invention is based, electronic device (2) Taking the raw signals collected by NPU, GPU or optimized 15 as input. (i) intent category (a predefined model running on a CPU) (ii) a label from the category space, (iii) the strength of the signal of interest (normalized value), It is structured to generate availability inferences. Artificial intelligence inference engine (3), inference result intention category, interest signal strength, 20 fields including availability, privacy label, TTL, alias, and device capability statement. packaging into a context token structure and determining the token size from the raw data. to adjust it to be smaller in terms of magnitude It is being structured. The token transmission module (4) in the system (1) which is the subject of the invention, artificial intelligence inference 25 the context marker generated and packaged by the engine (3) is pre-selected by the operator. 5G low latency in the form of a network slice or URLLC stream allocated by it. any guarantee of end-to-end 10 ms over communication infrastructure It is configured to be transmitted to the operator side within the allocated delay budget. This deterministic delay guarantees the availability of context that is "meaningful at that very moment". It allows for evaluation within its window. The multi-channel orchestration engine (5) in the system (1) of which the invention is located, operator context marker received on the side, last triggered channels, response rates and customer interaction history in the form of channel preferences; channel's real-time delay suitability, customer profile suitability, regulatory compliance, and channel cost 5 Channel-suitability score for each available interaction channel; time interval policies in the form of restrictions, channel quotas and rules to prevent duplicate communication by combining the rules and generating a normalized compliance score for each channel It is configured to allow the selection of one or more channels above the threshold. The reversible action module (6) in the system (1) of which the invention is located, selected Channel modules related to channel operator for channels: SMS, email, push, IVR- callback, to enable the transmission of the trigger request in the form of an outgoing call. It is structured. The undoable action module (6) executes each trigger. An instantaneous reverse operation definition is generated (e.g., SMS sent → "SMS recall 15"). "Notification"; IVR callback scheduled → "canceled" along with compensation recorded callback It is configured to be written to the retrievable queue. Retrievable action. If the module (6) responds from one of the customer triggered channels, context The token may be withdrawn if its TTL (Total Price) reaches its limit or if policy rules require its cancellation. The queue is operated in a last-in, first-out (LIFO) pattern, triggering other channel 20 It is structured to ensure that compensatory steps are taken for this. Back The action module (6) can be taken, in partial compensation mode, only the affected sub- to ensure the wrapping of the chains and the protection of successful sub-chains It is being structured. In the system in question (1), the server (7) receives the context marker, score calculations, channel selection decision, triggering, success / failure, compensation execution of each Score values ​​for the step, policy used, timestamps, alias reference To ensure that the decision, along with its reasoning, is recorded in the permanent audit log. It is configured accordingly. Server (7) 30 when a new interaction channel is added. only the channel's channel-suitability score definition and compensation definition are in the protocol. 6 notification, orchestration, score engine or transmission layer It is configured to ensure that it is not changed. Server (7), next generation IoT or a new type of electronic device (2) in the form of a connected vehicle interior surface device by adding a skills declaration to enable its inclusion in the system It is being structured. 5 Industrial Application of the Invention The invention in question is in an electronic device (2) (mobile device, smart device) on the customer side. IoT devices, connected in-vehicle interaction surfaces, etc.) use locally running artificial intelligence. 10 A structured context token generated by the inference engine (3) is used in 5G low delayed communication infrastructure (network slicing, URLLC, or similar mechanisms) that it is transmitted to an orchestration engine (5) on the operator side; context marker, interaction history and channel-suitability of the orchestration engine (5) by combining their scores from multiple customer interaction channels (SMS, email, push, 15 (IVR - callback, outgoing call, etc.) selects the appropriate one or ones; and the selected one channel triggers, along with compensation definitions generated at the time of execution, share a common By being added to the queue, the customer can respond or the context can change, allowing the customer to re-enter the chain. It is related to a system (1) from which it can be withdrawn. Around these basic concepts, the invention is very much about “An Orchestration System (1)”. It is possible to develop various applications, and the invention described herein It cannot be limited to examples; it is essentially as stated in the requests.

Claims

7 REQUESTS 1. On a mobile network operator's infrastructure, on a customer-side end device. context extracted locally can be applied to multiple customer interaction channels and 5 used to reflect this in the services - Device used / carried by the customer and its location, movement status, and location. the situation, recent application interactions, numerous sensor outputs at least one electronic device that enables signal collection (2), -operated locally on the customer’s electronic device (2) and the device (2) The signals on it include intention category, interest signal strength, and availability. a context including status, privacy label and validity lifecycle fields an edge AI inference engine that produces the token (3), -the context marker in question, raw customer signals device (2) boundaries by keeping it within, 5G low latency communications infrastructure (network slicing, a 15 allocated via URLLC or similar mechanisms) A token transmitted to the operator side within a deterministic delay budget. transmission module (4), -the context marker in question, customer interaction history, and each existing one. by combining it with the channel-suitability score for the customer interaction channel, 20 who choose one or more of at least two configurable channels for the election a multi-channel orchestration engine (5), - Each interaction message triggered from the selected channels, at the time of execution. a sequential process along with a generated reverse operation (compensating action) definition the queue is filled when the customer responds through another channel or context When the validity period of the token expires, the queue in question becomes last-entry-25 A reversal that reverses the first-interest order and takes compensatory steps. action module (6) and -electronic device (2), artificial intelligence inference engine (3), token transmission module (4), multi-channel orchestration engine (5) and retractable action The joint management of module (6) during a single customer interaction event 30 8 and ensures that the decision at each stage is written to a permanent audit log. an orchestration system (1) characterized by having a small server (7).

2. Any device in the form of a smartphone, smartwatch, IoT device, or connected in-vehicle surface. Claim 5 is characterized by an electronic device (2) which is a customer end device. A system like the one in 1 (1).

3. Instantaneous signals from a tracker, device sensor, or caller ID working on it. Location area, whether in motion, device charge / battery status, active The application category and data such as last interaction times are 10. to allow collection and storage within the device's boundaries Claim 1 is characterized by an electronic device (2) configured to be or A system like the one in 2 (1).

4. To receive the raw signals collected by the electronic device (2) as input, 15 (i) with a model that runs on an NPU, GPU or optimized CPU (ii) intention category (a label from a predefined category space), (ii) interest signal strength (normalized value), (iii) availability inferences characterized by an artificial intelligence inference engine (3) structured to produce a system like any of the above requests (1). 20 5. Inference results are based on intent category, strength of interest signal, availability, and privacy. a context containing the tag, TTL, alias, and device capability declaration fields packaging the token into a structure and sizing the token from the raw data. artificial 25 configured to be smaller at the level from the above requests characterized by the intelligence inference engine (3) a system like any other (1).

6. Context generated and packaged by the AI ​​inference engine (3) the token, a network slice or URLLC stream 30 pre-allocated by the operator end-to-end 10 over the 5G low-latency communication infrastructure 9 operator within any guaranteed delay budget in milliseconds with the token transmission module (4) configured to transmit to the side a system like any of the above characterized claims (1).

7. On the operator side, the received context marker is the last triggered. Customer interaction in the form of channels, response rates, and channel preferences. History; channel's real-time latency suitability, customer profile suitability, each existing interaction in the form of regulatory compliance and channel cost Channel suitability score for the channel; time slot restrictions, channel quotas and 10 policy rules in the form of rules to prevent duplicate communication by combining and generating a normalized fitness score for each channel, threshold configured to allow the selection of one or more channels at the top the above characterized by multi-channel orchestration engine (5) a system like any of the requests (1). 15 8. For selected channels, SMS, e- trigger requests in the form of mail, push, IVR callback, and outgoing call Reversible action module configured to ensure transmission (6) a 20 as in any of the above claims characterized by system (1).

9. Each trigger generates a reverse operation definition at the time of execution (e.g., SMS) Sent → "SMS recall notification"; IVR callback scheduled → "cancellation" along with compensation will be added to the refundable queue numbered 25. characterized by the configured revocable action module (6) a system like any of the above requests (1).

10. If the customer responds from one of the triggered channels, the context token If the TTL reaches its limit or policy rules require cancellation, it can be reversed. The queue is operated in a last-in, first-out (LIFO) configuration, with the other channel to ensure that compensatory steps are taken for the triggers characterized by the configured revocable action module (6) a system like any of the above requests (1).

11. In partial compensation mode, only the affected sub-chains are wound, successful 5 Reversible systems designed to ensure the protection of sub-chains. any of the above requests characterized by action module (6) a system like one of them (1).

12. Context marker acquisition, score calculations, channel selection decision, triggering, 10 Success / failure, score values ​​for each step of the compensation process, used policy, timestamps, pseudonym references, and justification for the decision configured to ensure that they are written to the persistent audit log together. any of the above requests characterized by the server (7) such a system (1). 15 13. When a new interaction channel is added, only the channel's channel-suitability is considered. Notification of score definition and compensation definition to the protocol, orchestration, score to ensure that the motor or transmission layer is not modified 20 of the above requests characterized by the configured server (7) a system like any other (1).

14. A new electronic device in the form of a next-generation IoT or connected in-vehicle surface. (2) Inclusion of type by adding device capability declaration to the system The above 25 is characterized by the server (7) configured to provide a system like any of the requests (1).

Images