Technical Overview

In order for the Alotech integration to work, Alotech Producer and Conversation Consumer services should be deployed to the desired environment.

On this page, it is explained how Knovvu Analytics works integrated with Alotech, the technical details and the required deployment steps in details

Alotech Services

1. Alotech Producer

One of the services written for Knovvu Analytics-Alotech integration is Alotech Producer. Alotech Producer's main task is to find all the calls starting from the requested date until today and transmit them to the consumer. For current calls, it keeps receiving them two hours behind.

Alotech Producer Deployment

Alotech Producer is designed to run separately for each tenant, so when deploying this service, it must be associated with a tenant through the information on the settings page.

When deploying Alotech Producer, the customer's tenant name should be entered as the variation name. The variant name should be the same as the tenant name, but completely in lowercase letters to avoid receiving errors due to the Regex rules. The number of replicas should be 1, so that the same event is not received more than once.

2. Conversation Consumer

The other service written is Conversation Consumer and its task is to receive the metadata and recordings of calls from RabbitMq via APIs and save them to Knovvu Analytics.

Conversation Consumer Deployment

Conversation Consumer is not tenant-based like Alotech Producer and is integration-type independent. This means that the same service can be used for different marketplace integrations as well as for all customers from different tenants. Therefore, the variation name of the Conversation Consumer is not important. And the number of replicas is increased according to the need.

When should a new Conversation Consumer deployed?
When the calls pile up in the RabbitMq queue, and cannot be processed, a new Conversation Consumer should be deployed. This is done by increasing the number of replicas.

Alotech Integrations Configuration Steps

In order for a customer to use the Alotech integration, this integration must be defined for the customer's tenant.

As the first step, the admin needs to login to the desired environment as a host and enable the Alotech feature for the tenants who will use this integration.

1. Login as host.
2. Enter the Tenants page under the Saas menu as seen below.

3. Next to the desired tenant's name, click on the Actions button and then Features.
4. Alotech Integration feature can be seen on the opened page and shoul be enabled here.

5. After these steps are completed on the admin side, you should login as the tenant.
6. Go to Sources page under Administration -> Definitions.
7. Create a new source by clicking on Add new.
   • Source name: Source name can be same as the tenant name.
   • Source type: Source type should b Http Media Service
   • Settings: Settings hould be:

{
  "Endpoint": "http://marketplace-media-server-marketplace.sestek-ca/integration/marketplace-media-server/marketplace/api/MarketPlaceMediaServer/alotech/{tenantId}/{externalId}"
}

8. Once the source is created, required integration settings should be filled. Enter the Settings page under the Administration menu. The information that needs to be entered for the Alotech integration to work can be seen below under the Alotech Integration tab.

   • Client ID: Customer's Alotech clientID/token.
   • Tenant Name: As the tenant name, the domain URL that the customer uses to access Alotech must be entered. As an example, the tenant xyz who accesses the alotech environment from the xyz.alo-tech.com url should write xyz.alo-tech as the tenant name.
   • Source Name: The name of the source created in the 7th step should be entered here.
   • Migration Start Date: The date from which the calls will be analyzed.
     Although the service has been launched today, calls from 1 year ago can be analyzed if desired.
   • Default Language Code: If a new language model developed specifically for the customer and not predefined in the system wants to be used, the code of the language model must be entered in this field. It is not a mandatory setting, and can be left blank when not used.
   • Queue Filter: If queue-based filtering wants to be applied instead of analyzing all customer calls, this checkbox should be selected. Queue Filter Dictionary is the field used to define queue-based language code. Alotech does not send language code information. Therefore, the analysis language must be entered manually. Information about which language code will be used for each queue to be listened, to should be defined in this field. {"Queue ID": "language code"} format must be applied.
   • Agent Filter: If agent-based filtering wants to be applied instead of analyzing all agents' calls, this checkbox should be selected. The *IDs or names of the agents, whose calls will be analyzed should be defined in this field with a comma between them.
   • Custom Field Filter: If custom filtering will be done based on any other parameter, this checkbox should be selected. The parameter to be filtered and the value it should take should be defined in this field. If more than one parameter is defined, only the calls, for which all filters are met will be analyzed, meaning there is an and relationship between custom filters, not or.

Once these steps are completed, the admin needs to login as host and assign the service to the tenant.
9. Login as host.
10. Enter the Integration page under the Administration menu as seen below.

11. All services in the environment can be seen on this page. By clicking on the AlotechProducer, the user must enter the service whose status is waiting for configuration. The tenant ID of the customer should be entered.

12. After clicking on Save, the waiting for configuration status of the service will be switched to running. Alotech integration is ready to be used by the customer.