Multi-node simultaneous concurrent calling method for intelligent customer service calling robot

Abstract

The invention discloses a multi-node simultaneous concurrent calling method for an intelligent customer service calling robot, and belongs to the technical field of customer service calling control, and the method comprises the following steps: S1, building a node cluster service according to each existing calling node, and monitoring the calling state and load rate of each node after the node cluster service is connected with each node; and S2, after the call task starts, the node cluster service obtains the current call volume and the corresponding load rate of each node, and distributes the call task to the node of which the current load rate meets a preset condition. According to the method, a calling strategy based on load balancing is adopted, the calling task of the original service is independent of one node cluster service, the calling tasks of multiple nodes are distributed, the method of multi-node simultaneous concurrent calling is achieved, the calling tasks are reasonably distributed to all the nodes according to the load rates of all the nodes before calling, and therefore the calling efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of customer service call control, and in particular to a method for simultaneously making multiple concurrent calls by an intelligent customer service call robot. Background Art

[0002] With the continuous development of science and technology, artificial intelligence has gradually penetrated into all walks of life. The emergence of intelligent call robots has brought a new customer experience to operators and enterprises. Compared with traditional manual calls, intelligent call robots have the advantages of higher efficiency, lower cost and better user experience. Traditional call robots generally use a single-node call model to perform call tasks.

[0003] The traditional single-node call model has a large node load rate, low efficiency, insufficient stability and scalability when the call task volume is large, and as the call task volume expands, a single node cannot meet business needs. The above problems need to be solved urgently. To this end, a multi-node concurrent calling method for an intelligent customer service call robot is proposed. Summary of the invention

[0004] The technical problem to be solved by the present invention is: how to solve the problems of poor stability, low throughput and poor scalability of the existing single-node calling method, and provide a multi-node concurrent calling method for an intelligent customer service calling robot. The method adopts a call strategy based on load balancing, which makes the call task of the original service independent as a node cluster service, distributes the call tasks of multiple nodes, and realizes a method for multi-node concurrent calling. Before the call, the call tasks are reasonably allocated to each node according to the load rate of each node, thereby improving the calling efficiency.

[0005] The present invention solves the above technical problems through the following technical solutions, and the present invention comprises the following steps:

[0006] S1: Establish ESL cluster service based on existing call nodes, and after connecting with each call node, feedback the current connection status of each call node;

[0007] S2: After the call task starts, the call data is pushed from the cache queue to the node cluster service at a regular interval. After the node cluster service obtains the call data, it first obtains the current call volume and corresponding load rate of each call node, and distributes the call data to the call node whose current load rate meets the preset conditions to execute the call.

[0008] Furthermore, in step S1, the specific process of the ESL cluster service establishing a connection with each calling node is as follows:

[0009] S11: Write configuration information of each call node into the ESL cluster service configuration file;

[0010] S12: Start the ESL cluster service, establish a unified node management thread pool, and each node thread service reads the cluster service configuration file information to establish a connection with each calling node;

[0011] S13: The calling nodes with failed connection are removed, and the node connection success or fault node information is printed and fed back in the form of logs.

[0012] Furthermore, in step S2, the specific process of obtaining the current call volume of each call node and the corresponding load rate is as follows:

[0013] S21: The node cluster service first obtains the current status of each calling node, including the current call volume of the node and the maximum call volume of the node, wherein the current call volume of the node is the number of ongoing calls of the current calling node, and the maximum call volume of the node is the maximum number of calls supported by the current calling node;

[0014] S22: Calculate the load rate according to the current call volume of the node and the maximum call volume of the node. If there is a faulty node, it is directly removed and does not participate in the load rate calculation.

[0015] Furthermore, in step S22, load rate = current call volume of the node / maximum call volume of the node.

[0016] Furthermore, in step S2, the specific process of dispatching the waiting-for-call data to the calling node whose current load rate meets the preset conditions to execute the call is as follows:

[0017] S23: After obtaining the load rate of each call node, sort the load rates to obtain the nodes with lower load rates;

[0018] S24: Each piece of call data itself has executable script command parameters. After obtaining the node with a lower load rate, the ESL cluster service sends batch execution script commands to the call node based on the obtained outbound call data. The call node accepts the execution request, parses the script command and executes the specific call process.

[0019] Furthermore, in step S23, when the load rates of the call nodes are sorted, the nodes with lower load rates are the call nodes whose load rates are less than a set threshold.

[0020] Furthermore, in step S2, for a faulty node, the node cluster service excludes the node when dispatching a call task, and performs log feedback at the first time, and will continue to feedback the status of the node when dispatching a call task next time.

[0021] Furthermore, the faulty node is determined as follows: when the ESL cluster service is started, a call node fault detection is performed, and the node can be determined to be a faulty node based on the node error information; after the ESL cluster service receives the distributed call data, it will obtain the load rate of each node in real time. If the load rate of a node cannot be obtained correctly, it can be determined to be a faulty node.

[0022] Compared with the existing technology, the method has the following advantages: the method for concurrent calling of multiple nodes of the intelligent customer service call robot first establishes the connection between the node cluster service and the existing nodes, and monitors the node status, node call volume and node load rate of each node in real time; then, when the call task starts, the available nodes are first screened out, and then the call tasks are distributed to each node more evenly according to the number of calls and the current load rate of each available node; finally, for the nodes that fail and are unavailable, log feedback and continuous monitoring feedback are performed; it can greatly improve the utilization rate and stability of each node, and at the same time achieve better horizontal expansion capabilities, and call nodes can be added later according to the expansion of business. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flowchart of a method for simultaneously calling multiple nodes of an intelligent customer service calling robot in Embodiment 1 of the present invention;

[0024] Figure 2 It is a flowchart of the method for simultaneously calling multiple nodes of the intelligent customer service calling robot in the second embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0026] Embodiment 1

[0027] like Figure 1 As shown, this embodiment provides a technical solution: a method for simultaneously calling multiple nodes of an intelligent customer service calling robot, including the following main processes:

[0028] Step 1: Establish ESL cluster service based on existing call nodes, and after connecting to each call node, feedback the current connection status of each call node.

[0029] In this step, the specific process of the ESL cluster service establishing a connection with each calling node is as follows:

[0030] 1): Write the configuration information of each call node in the ESL cluster service configuration file;

[0031] 2): Start the ESL cluster service, establish a unified node management thread pool, and each node thread service reads the cluster service configuration file information to establish a connection with each calling node;

[0032] 3): The calling nodes that failed to connect are removed, and the node connection success or fault node information is printed and fed back in the form of logs.

[0033] Step 2: After the call task starts, the call data is pushed from the cache queue to the node cluster service at a regular interval. After the node cluster service obtains the call data, it first obtains the current call volume and corresponding load rate of each call node, and distributes the call data to the call node whose current load rate meets the preset conditions to execute the call.

[0034] The specific process of obtaining the current call volume and corresponding load rate of each call node is as follows:

[0035] 1): The node cluster service first obtains the current status of each calling node, including the number of calls currently in progress and the maximum number of calls supported, that is, the current call volume of the node and the maximum call volume of the node;

[0036] 2): (current call volume of each node / maximum call volume of each node) is the load rate of each node;

[0037] 3): If there is a faulty node, it will be directly eliminated and will not participate in the load condition calculation.

[0038] In this step, the specific process of dispatching the waiting-for-call data to the calling node whose current load rate meets the preset conditions to execute the call is as follows:

[0039] 1): After obtaining the load rate of each node, a load rate ranking will be performed, that is, the node with the lowest load rate will be ranked;

[0040] 2): Distribution process: Each piece of call data itself has executable script command parameters. After obtaining the node with a lower load rate, the LUA script command can be sent in batches to the call node through the ESL cluster service according to the obtained outbound call data to execute the outbound call request.

[0041] 3): The cluster load node accepts the execution request, parses the script command and executes the specific call process.

[0042] In step 1), the node with a lower load rate is a calling node whose load rate is less than a set threshold.

[0043] Step 3: For a faulty node, the node cluster service excludes the node when dispatching a call task, and provides log feedback immediately. The node status will continue to be fed back when the call task is dispatched next time.

[0044] In this step, when the ESL cluster service is started, a call node fault detection is performed, and the node can be determined to be a faulty node based on the node error information; after the ESL cluster service receives the distributed call data, it will obtain the load rate of each node in real time. If the load rate of a node cannot be correctly obtained, it can be determined to be a faulty node.

[0045] In this embodiment, in some actual business scenarios, the concurrency capacity of intelligent outbound calls is improved by horizontal expansion capabilities. If you want to expand the outbound capacity by adding call nodes, you can add the configuration information of the call nodes in the ESL cluster service, and then restart the ESL cluster service. The ESL cluster service automatically parses and loads to expand the outbound call capacity and realize the horizontal expansion of outbound calls.

[0046] Embodiment 2

[0047] like Figure 2 As shown, this embodiment further illustrates the method for simultaneously calling multiple nodes of the intelligent customer service calling robot in Embodiment 1, and specifically includes the following process:

[0048] 1. Node cluster service deployment:

[0049] First, configure the existing resource nodes in a unified manner to facilitate unified management of the node cluster service. Then configure the parameters of each node in the node cluster service configuration file. Finally, after starting the node cluster service, you can check the current status and load rate of each node through scheduled tasks, and also scan for call tasks.

[0050] 2. Call task execution:

[0051] After starting the call task, the application service will store the call task. The scheduled task in the node cluster service will monitor the status of each node, scan the call tasks stored in the application service and extract them. Finally, according to the status and load rate of each node, the call task will be dispatched, so as to maximize the utilization of each call node and ensure the high efficiency of each call node.

[0052] To sum up, the multi-node concurrent calling method of the intelligent customer service call robot in the above embodiment first establishes a connection between the node cluster service and the existing nodes, and monitors the node status, node call volume and node load rate of each node in real time; then, when the call task starts, the available nodes are first screened out, and then the call tasks are distributed to each node more evenly according to the number of calls and the current load rate of each available node; finally, for the nodes that fail and are unavailable, log feedback and continuous monitoring feedback are performed; it can greatly improve the utilization and stability of each node, and at the same time achieve better horizontal expansion capabilities, and call nodes can be added later according to the expansion of business.

[0053] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for simultaneously calling multiple nodes of an intelligent customer service calling robot, characterized in that: The following steps are involved: S1: Establish ESL cluster service based on existing call nodes, and after connecting with each call node, feedback the current connection status of each call node; S2: After the call task starts, the call data is pushed from the cache queue to the node cluster service at a regular interval. After the node cluster service obtains the call data, it first obtains the current call volume and corresponding load rate of each call node, and distributes the call data to the call node whose current load rate meets the preset conditions to execute the call.

2. According to claim 1, a method for simultaneously calling multiple nodes of an intelligent customer service calling robot is characterized in that: In step S1, the specific process of the ESL cluster service establishing a connection with each calling node is as follows: S11: Write configuration information of each call node into the ESL cluster service configuration file; S12: Start the ESL cluster service, establish a unified node management thread pool, and each node thread service reads the cluster service configuration file information to establish a connection with each calling node; S13: The calling nodes with failed connection are removed, and the node connection success or fault node information is printed and fed back in the form of logs.

3. According to claim 2, a method for simultaneously calling multiple nodes of an intelligent customer service calling robot is characterized in that: In step S2, the specific process of obtaining the current call volume of each call node and the corresponding load rate is as follows: S21: The node cluster service first obtains the current status of each calling node, including the current call volume of the node and the maximum call volume of the node, wherein the current call volume of the node is the number of ongoing calls of the current calling node, and the maximum call volume of the node is the maximum number of calls supported by the current calling node; S22: Calculate the load rate according to the current call volume of the node and the maximum call volume of the node. If there is a faulty node, it is directly removed and does not participate in the load rate calculation.

4. According to claim 1, a method for simultaneously calling multiple nodes of an intelligent customer service calling robot is characterized in that: In the step S22, load rate = current call volume of the node / maximum call volume of the node.

5. According to claim 3, a method for simultaneously calling multiple nodes of an intelligent customer service calling robot is characterized in that: In step S2, the specific process of dispatching the waiting-for-call data to the calling node whose current load rate meets the preset conditions to execute the call is as follows: S23: After obtaining the load rate of each call node, sort the load rates to obtain the nodes with lower load rates; S24: Each piece of call data itself has executable script command parameters. After obtaining the node with a lower load rate, the ESL cluster service sends batch execution script commands to the call node based on the obtained outbound call data. The call node accepts the execution request, parses the script command and executes the specific call process.

6. According to claim 4, a method for simultaneously calling multiple nodes of an intelligent customer service calling robot is characterized in that: In the step S23, when the load rates of the call nodes are sorted, the nodes with lower load rates are the call nodes with load rates less than the set threshold.

7. The method for simultaneously calling multiple nodes of an intelligent customer service calling robot according to claim 6 is characterized in that: In step S2, for a faulty node, the node cluster service excludes the node when dispatching a call task, and performs log feedback immediately, and will continue to feedback the status of the node when dispatching a call task next time.

8. The method for simultaneously calling multiple nodes of an intelligent customer service calling robot according to claim 7, characterized in that: The method for determining a faulty node is as follows: When the ESL cluster service is started, a call node fault detection is performed, and a faulty node can be determined based on the node error information; after the ESL cluster service receives the distributed call data, it will obtain the load rate of each node in real time. If the load rate of a node cannot be obtained correctly, it can be determined as a faulty node.

Images