Connector implementation checklist

This guide covers the tasks that a client administrator must complete to configure connectors.

Important: Some of the following tasks must be completed before a Medallia administrator configures Experience Cloud to allow the connector to interact with it. For this reason, Medallia administrators cannot help client administrators troubleshoot these steps.

NICE inContact (Central)

The Medallia Speech connector for NICE inContact integrates via APIs with NICE inContact (both CXOne Central and UserHub). If you are a customer of other NICE systems, contact your Medallia expert to manage accessibility.

Start by creating a dedicated security profile for Medallia Speech or identifying one with the necessary access:

  • Name — Medallia Speech Connector

  • Contact history — Agent View + Supervisor View

  • Data download — View

Next, create an agent account for Medallia Speech:

  • Vendor name — Medallia

  • Application name — MedalliaExperienceCloud

  • Application ID — Provide an application ID.

  • Agent username — Provide a username.

  • Agent password — Enter a password.

  • Security profile — Medallia Speech Connector (or equivalent)

Enter the application name value without spaces between the words. Camel case is recommended, as shown above, to help with readability.

Once the agent account has been created, you must change the initial password by logging in to https://login.incontact.com in order to activate the agent account for API access. After this, provide the changed password to Medallia.

NICE inContact (UserHub)

Before configuring your connector in Medallia Experience Cloud, you must configure your NICE system for inbound communication. Once that is done, you can complete the connector configuration.

Start by creating a dedicated client for Medallia Speech or identifying one with the necessary access:

  • CompletedContact API — Read

    NICE inContact UserHub chat and email connectors also use this API. For information about the API, see the NICE inContact Reporting APIs documentation.

  • MediaPlayback API — Read

  • Files API — Read

  • API domain (region/data center) — Values include: na1 (North America), eu1 (Europe), and au1 (Australia).

    UserHub uses different servers for authentication and the UserHub API, so you should first determine the region in which your instance of NICE inContact UserHub runs. The selected region will determine the authorization server and the API server used.

Application requests that are not made through a browser require an access key. Access keys consist of two keys:

  • Access Key ID — Provide the key ID.

  • Secret Access key — Provide the secret access key.

Use the Access Key ID and Secret Access Key together to authenticate your client requests.

Genesys Cloud (PureCloud)

The Medallia Speech connector for Genesys Cloud integrates via APIs with Genesys Cloud Contact Center. If you are a customer of other Genesys systems, contact your Medallia expert to manage accessibility.

Before configuring your connector in Medallia Experience Cloud, the client administrator must configure their Genesys Cloud system for inbound communication. Once that is done, the administrator will provide the information that you will need to complete the connector configuration.

Start by creating a dedicated user for Medallia Speech or identifying one with the necessary permissions:

required permissions
  • Recording > Recording > View

  • Recording > Annotation > View

  • Analytics > Conversation Detail > View

  • Routing > Queue > View

  • Routing > Skill > View

Note:

The Routing > Queue > View and Routing > Skill > View permissions allow the connector to verify data mapping by sending requests to the following endpoints:

  • https://api.<region>.pure.cloud/api/v2/routing/queues
  • https://api.<region>.pure.cloud/api/v2/routing/skills

Determine the region in which your instance of Genesys Cloud runs, so that your application makes calls to that region.

  • API region — Examples include US East Virginia, US West Oregon, Europe London, and Asia Pacific Sydney.

Next, create a client application (OAuth client) for Medallia Speech with the necessary access:

  1. In Genesys Cloud, create a new OAuth client using the Client Credentials grant type.

  2. Enter "MedalliaExperienceCloud" in the App Name field.

  3. Copy the generated Client ID and Client Secret to a secure location so that you can use them to authenticate the integration.

  4. In the Roles tab, assign a custom role that contains the necessary view permissions for recordings, conversations, and analytics:

    required permissions

    • Recording > Recording > View

    • Conversation > Communication > View

    • Analytics > ConversationDetail > View

    • Analytics > ConversationAggregate > View

  5. Save the configuration when finished.

Genesys Cloud APIs use the OAuth 2.0 protocol for authentication. Use the Client ID and Client Secret together to send an authorization request to the authorization server so that it returns an access token for your client application.

Note: Genesys Cloud has different servers for authentication and for the Cloud API.

Considerations for splitting interaction segments

The Genesys Cloud connector supports splitting an interaction into multiple segments to manage calls with multiple units. If you would like to split segments in Experience Cloud, enable the Split Segments setting when configuring the connector.

For more information about how this works in Experience Cloud, see Connector settings and Multi-touch programs.

RingCentral Engage

RingCentral Engage is a contact center as a service offering.

Warning: RingCentral Engage APIs are limited to production environments only, not sandboxes. Testing the connector can only occur in a production environment of RingCentral Engage.

There are two main, different products from RingCentral: RingCentral Office and RingCentral Engage. The Medallia Experience Cloud connector only supports RingCentral Engage.

RingCentral Engage is a contact-center-as-a-service offering. It is an acquired product, yet it is integrated with the core RingCentral platform. However, there is still a limitation with RingCentral Engage: their APIs are limited to production environments only, not sandboxes. So testing the connector can only occur in a production environment of RingCentral Engage.

Medallia Experience Cloud connector works with RingCentral Contact Center, also known as Engage, APIs. However, you will need to configure settings in both RingCentral MPV and RingCentral Contact Center in order to properly setup the connector.

Note that:

  • There are no sandbox environments for RingCentral Contact Center (formerly known as RingCentral Engage), there is only a production instance. RingCentral MVP offers clients a sandbox environment, but it cannot be used with RingCentral Contact Center.

  • RingCentral MVP and RingCentral Contact Center have separate administration portals:

  • You must create App and API credentials in RingCentral MVP. While the connector will work with RingCentral Engage APIs, the credentials (client ID and client secret) are created and managed in RingCentral MVP (RingCentral Office).

  • You need both, API credentials from RingCentral MVP and user credentials from RingCentral Contact Center to configure the Medallia RingCentral Speech connector.

Requirements

First, set up RingCentral to record calls in stereo. This ensures that Experience Cloud can distinguish between agents and customers.

To setup the Medallia Speech Connector for RingCentral Contact Center, collect the following information:

  • App and API credentials for RingCentral MVP production environment:

    • Client ID

    • Client secret

  • User Credentials for a RingCentral Contact Center user-assigned a role with the requisite permissions.

    • User credentials are:

      • Company (RingCentral MVP) main phone number

      • User’s RingCentral MPV office extension

      • User’s password

    • The user must have, at least, the following permissions:

      • AccountView

      • AccountReport

API credentials

Access your RingCentral MVP portal. This is typically located at http://service.ringcentral.com.

To create a set of API credentials on the RingCentral MVP portal:

  1. Go to the developer portal.

    Screen capture showing RingCentral home page with selected Developer Portal option

  2. Go to the MVP console.

    Screen capture of MVP console with highlighted Create App option

  3. Select Create App.

  4. Select REST API App and click Next.

    Screen capture of create app form with highlighted REST API app option

  5. In the settings, configure the following properties and then click Create.

    • App Name — (Internal-use only) Enter a descriptive name, such as “My Medallia RingCentral Connector App”.

    • Display Name — (what others will see) Enter a descriptive name, such as “My Medallia RingCentral Connector App”.

    • App Description — Enter a brief description, such as “This app connects RingCentral Engage with Medallia Speech”.

    • App Icon — There is no need to upload an icon.

    • Primary Contact — Set this to a valid RingCentral user.

    • App Promotion — As this is a private app, select No when asked if you plan to promote this app.

      Screen capture showing RingCentral app settings

    • Auth — Select password-based authentication flow.

    • App Permissions — Grant the following permissions:

      • Read Call Recording

      • Read Call Log

    • App Access — Select the option This app is private and will only be callable using credentials from the same RingCentral account.

    • Interactive Messages — Leave unchecked.

      Screen capture showing RingCentral app settings

Once you create your app, it is ready for use in a sandbox environment and needs to be promoted to production. As RingCentral Engage does not offer a sandbox environment, work with either your RingCentral account liaison or Medallia expert to have your app promoted by RingCentral.

Once your app is promoted, you will see a Client Id and Client Secret for production. You will need that in addition to the account phone number.

Screen capture showing promoted app client ID and client secret

User credentials

Next, create a user within both RingCentral MVP and RingCentral Engage.

To create a free user account:

  1. On the RingCentral MVP portal, select Add User.

    Screen capture of RingCentral MVP portal with highlighted "Add User" option

  2. Select Mobile User, Free, and click Select.

    Screen capture of user-type selection form with highlighted "Mobile user" and "Free" options

  3. Enter the user’s first name, last name, email address, and make note of the RingCentral MVP extension that is assigned to this user.

    The current best practice is to assign a descriptive name to the user such as "Medallia Connector".

    Screen capture of "Add user info" pane with highlighted "Extension" text box

  4. Click Next.

  5. Select Activate by assigning credentials, and configure the password, PIN, security question and role.

    The RingCentral predefined Standard (International) role is sufficient for this connector.

    Screen capture of user setup options pane

Once the user account has been created in RingCentral MVP, it must then be synced to RingCentral Contact Center (Engage).

To sync the user account:

  1. On the RingCentral Contact Center portal, select Admin.

    Screen capture of the RingCentral Contact Center portal home screen with highlighted Admin option

  2. On the Users section, select Administrators.

    Screen capture of Engage Voice Admin screen with highlighted Administrators option

  3. Select New User.

    Screen capture of administrators setup page with highlighted New user option

    On the New user screen, do not enter any information such as first name, last name, etc.

  4. Expand the RC Office extension dropdown list. There is a list of users from RingCentral MVP (also known as "RingCentral Office") that do not have a corresponding account within RingCentral Engage.

  5. Select the account you created from that list.

    The rest of the fields are automatically populated.

  6. On the Basic info tab, configure the following property:

    • Access Historical AnalyticsViewer

      Screen capture of the user basic info tab with highlighted Access Historical Analytics option

  7. Click Save.

  8. Next, assign a role to this user with the following minimum permissions:

    • AccountView

    • AccountReport

Once completed, gather the following information:

  • App and client ID

  • App and client secret

  • User phone number, extension, and password

With this information, contact your Medallia expert to configure the Medallia Speech RingCentral Connector in your Medallia Experience Cloud instance.

Five9

The Medallia Speech connector for Five9 currently integrates with Five9 Cloud Call Center. If you are a customer of other Five9 systems, contact your Medallia expert to manage accessibility.

Call metadata are transferred over SFTP, typically in hourly batches.

To increase the effectiveness of the Medallia analysis, call recordings are recommended to be stitched together based on the Call.session_id value, and only stitched recordings should be exported in the Call Log reports and uploaded to the SFTP server.

Call Log reports must be exported into a CSV format, with Windows-formatted (CRLF) line breaks. The reports contain the call Call.session_id and Call.talk_time data points.

As part of the connector, call recordings are associated with their Call Log entries, so the call recording filenames must be based on data available in the Call Log reports. It is recommended that the format of the recording filename be Call.session_id.wav.

Call recordings are transferred over SFTP.

Medallia Media File Transfer

Due to character limitations on the SFTP username and password fields allowed by Five9 Cloud Call Center, the default SFTP account created during Speech provisioning requires modification when you use Five9. File a support ticket at https://help.medallia.com to request the addition of a Five9-compatible SFTP account to your existing Medallia Speech storage location. For information about MMFT, see Medallia Media File Transfer screen.

Connector setup

To set up the Medallia Speech connector, configure both call log reports and call recordings to be delivered to an SFTP location in separate folders. Once you complete these steps, provide the SFTP access information to your Medallia expert or contact your Medallia team if you have any questions.

Audio recordings

In the VCC Configuration, select the Advanced Recordings tab.

Create a new Destination:

  • NameMedallia Speech Audio Recording

  • Protocol — SFTP

  • Hostnamefilestash.{DC}.medallia.{TLD}

    Tip: Enter valid "DC" and "TLD" values for your installation.
  • Port — 22

  • Username — Enter a username for your installation.

  • Password — Enter a password for your installation.

  • Prefix — Leave blank

  • Time Zone — Default

  • Max Connections — 30

  • Rewrite existing files — Checked

Next, create a new Upload Rule:

  • NameStitched Recordings for Medallia Speech

  • Filter

    • Condition Grouping — Define appropriate conditions for your business logic.

    • Expression — Enter an appropriate expression for your business logic.

  • Destinations

    • DestinationMedallia Speech Audio - Stitched Recording

    • Filter Expression CriteriaCombined Recording - Match Filter

    • Recording file name pattern

      • Constantmedallia-speech-recordings

      • Directory delimiter/

      • Call variable{Call.session_id}

      • Constant.wav

The file name preview should read: medallia-speech-recordings/CallSessionId.wav.

Metadata

Start by creating a Scheduled Report:

  • Format — Tabular
  • Columns
    • Time
      • DATE
      • TIMESTAMP
    • Recording
      • DELETED
      • LENGTH
    • Recording Upload
      • DESTINATION
      • ERROR MESSAGE
      • FILE PATH
      • FILENAME
      • FILESIZE
      • HOSTNAME
      • STATUS
      • STITCHED
      • UPLOAD ATTEMPTS
      • UPLOAD TIMESTAMP
    • Calls
      • ANI
      • CALL ID
      • CAMPAIGN
      • DNIS
      • SESSION ID
      • SKILL
    • Call Statistics
      • HANDLE TIME
      • HOLD TIME
      • HOLDS
      • QUEUE WAIT TIME
      • TALK TIME
    • Agent
      • AGENT
      • AGENT_ID
      • AGENT_NAME
  • Report Ordering — Medallia recommends that you have SESSION ID, DATE, and TIMESTAMP at the top.
  • Summary Fields — Ignore. Summary fields do not apply to tabular reports.
  • Order information — SESSION ID then DATE, then TIMESTAMP.
  • Report Options
    • Time Frame Interval — Last hour (to match the schedule).
    • Advanced Filters — Define valid advanced filters for your business logic and needs.

Next, create a Scheduled Report Export:

  • Scheduled Report NameMedallia Speech Call Metadata (Hourly)
  • Scheduled Report Descriptionmail2ftp: {transport: 'SFTP', host: 'filestash.DC.medallia.TLD', port: '22', login: 'USERNAME', password: 'PASSWORD', folder: 'medallia-speech-metadata', filename: "metadata_%Y%m%d%H%M.csv", onFailure: 'email1@domain'}
    Tip: Enter valid "DC", "TLD", "USERNAME", "PASSWORD", and "onFailure" values for your installation.
  • Delivery Report — via E-mails
  • E-mailsemail1@domain
    Tip: Enter a valid email address for your installation.
  • Time Zone — Select valid timezone for your installation.
  • Output Format: — CSV (Windows)
  • Hourly — Select all hours shown.
  • Minutes — 00
  • Shared — Checked
  • Enabled — Checked

SFTP

Call recordings are transferred over SFTP.

To increase the effectiveness of the Medallia analysis, call recordings are recommended to be stitched together based on the Call.session_id value, and only stitched recordings should be exported in the Call Log reports and uploaded to the SFTP server.

Call Log reports must be exported into a CSV format, with Windows-formatted (CRLF) line breaks. The reports contain the call Call.session_id and Call.talk_time data points.

As part of the connector, call recordings are associated with their Call Log entries, so the call recording filenames must be based on data available in the Call Log reports. It is recommended that the format of the recording filename be Call.session_id.wav.

Connector setup

To set up the Medallia Speech connector, configure both call log reports and call recordings to be delivered to an SFTP location in separate folders. Once completed, provide the SFTP access information to your Medallia expert or contact your Medallia team if you have any questions.

Start by creating the call recording exports in the administrator application with the given upload rule:
  • NameStitched Recordings for Medallia Speech

  • Filter expression criteria — Combined Recording - Match Filter

  • Destination — Stitched Recordings

Next, modify the edit destination and file name pattern:
  • Destination — Stitched Recordings

  • Delivery — Combined Recording - Match Filter

  • Recording file name pattern

    • Constantmedallia-speech-recordings

    • Directory delimiter/

    • Call variable{Call.session_id}

    • Constant.wav

The file name preview should read: medallia-speech-recordings/CallSessionId.wav.

S3 bucket

The Medallia Speech connector for S3 buckets brings data into Medallia Experience Cloud by retrieving files from an S3 bucket. Connectors run on a schedule and access the S3 bucket via a file transfer protocol (FTP).

To increase the effectiveness of the Medallia analysis, call recordings are recommended to be stitched together based on the Call.session_id value, and only stitched recordings should be exported in the Call Log reports and uploaded to the SFTP server.

Call Log reports must be exported into a CSV format, with Windows-formatted (CRLF) line breaks. The reports contain the call Call.session_id and Call.talk_time data points.

As part of the connector, call recordings are associated with their Call Log entries, so the call recording filenames must be based on data available in the Call Log reports. It is recommended that the format of the recording filename be Call.session_id.wav.

For each call record, the following information must be imported: audio file name, call identifier, unit identifier, and response date.

Connector setup

To set up the Medallia Speech connector, configure both call log reports and call recordings to be delivered to an S3 location in separate folders. Medallia administrators need Read access to the directories and files. Once completed, provide the access information to your Medallia expert or contact your Medallia team if you have any questions.

Medallia administrators connect to the bucket using the standard S3 login credentials, all of which must be provided to Medallia. For information about these credentials see S3 Bucket and Amazon Connect.

Amazon Connect

Medallia Experience Cloud connectors for Amazon Connect are not functionally different to the generic S3 connectors. The key difference is that Amazon Connect supports chat ingestion along with audio. Generic S3 connectors do not support chat.

Some of the connector setup requires an Amazon Connect administrator on the client side to set up the necessary data streams to transfer audio and metadata to an S3 bucket.

Experience Cloud picks up the data from the prescribed S3 bucket that an Amazon administrator sends to Medallia. The connector setup is almost identical to the setup of the generic S3 connector.

Then, for Amazon Connect this connector type requires an AWS access account.

On the Amazon Connect console:

  1. Set up Amazon Connect to store call recordings and chat transcripts:

    1. Go to Instances and select the virtual contact center instance you want to connect to.

      Screen capture showing the Instances section of the Amazon Connect console

    2. On the Virtual Contact Center instance selected, go to Data Storage and verify that the Call recordings tile is showing a S3 bucket link, check the same thing for the Chat Transcripts tile.

      Screen capture showing the Data Storage section with the Call recordings and Chat transcripts tiles

    3. Click Edit on the Call recordings tile.

      Screen capture showing the expanded Call recordings tile

    4. Make sure the Enable call recordings option is selected.

    5. Select an existing bucket from the dropdown list or map a new bucket.

    6. Click Save.

      Screen capture showing S3 bucket options on Call recordings tile

    7. Click Edit on the Chat transcripts tile.

    8. Verify that the Enable chat transcripts option is selected.

    9. Select an existing bucket from the dropdown list or map a new bucket.

    10. Click Save.

      Screen capture showing S3 bucket options on Chat transcripts tile

  2. Create a Kinesis Data Stream to read from the bucket:

    1. On the Virtual Contact Center instance selected, go to Data Streaming and verify that the Enable data streaming option is selected.

      Screen capture showing the Data streaming section

  3. Create Kinesis Firehose using that data stream.

    1. On the Data Streaming tile, click Create a new Kinesis Firehose.

      Screen capture showing the create a new Kinesis Firehose option

      A new browser tab opens.

    2. On the new tab, select Kinetic Data Firehose.

      Screen capture showing the delivery streams creation options

    3. Click Create Delivery Stream.

      Delivery Stream information pane

  4. Enable Amazon Connect Data Streaming using Firehose.

    1. Specify the source and destination for your delivery stream.

      Screen capture showing the delivery stream info tiles with source and destination data

    2. Click Save.

Medallia administrators connect to the bucket using the standard S3 login credentials, all of which must be provided to Medallia. For information about these credentials see S3 Bucket and Amazon Connect.

Speech data flow

The diagram below shows the data flow for Amazon Connect speech connectors.

Diagram showing the data flow for Amazon connect speech connectors

Chat data flow

The diagram below shows the data flow for Amazon Connect chat connectors.

Diagram showing the data flow for Amazon connect chat connectors

Speech recording quality

Medallia Speech works best on high-quality voice recordings. Medallia recommends that all recordings be made with the highest fidelity possible and with each speaker given its own audio channel in the recording file. For that reason, recordings at a 44.1 kHz sample rate using multi-channel (ex: "stereo") 16-bit samples are strongly recommended.