Enable Contact Lens for Amazon Connect - Amazon Connect

Enable Contact Lens for Amazon Connect

You can enable Contact Lens for Amazon Connect in a few steps. Just add a Set recording and analytics behavior block to a flow, and configure it to enable Contact Lens for voice, chat, or both.

The following image shows a block that's configured for call recording, and Contact Lens speech analytics and chat analytics.

Collect data after transferring a contact: If you want to continue using Contact Lens to collect data after transferring a contact to another agent or queue, you need to add another Set recording and analytics behavior block with Enable analytics enabled for the flow. This is because a transfer generates a second contact ID and contact record. Contact Lens needs to run on that contact record as well.

Important

When you select Enable Contact Lens conversational analytics, you must choose to enable speech or chat analytics. Otherwise, your flow will display an error when you publish it.

The following procedures describe the steps to enable Contact Lens for call or chat analytics.

Enable call recording and speech analytics

  1. In the Set recording and analytics behavior block, under Call recording, choose On, Agent and Customer.

    Both agent and customer call recordings are required to use Contact Lens for voice contacts.

  2. Under Analytics, choose Enable Contact Lens conversational analytics, Enable speech analytics.

    If you don't see this option, Contact Lens for Amazon Connect hasn't been enabled for your instance. To enable it, see Update instance settings.

  3. Choose one of the following:

    1. Post-call analytics: Contact Lens analyzes the call recording after the conversation and After Contact Work (ACW) is complete. This option provides the best transcription accuracy.

    2. Real-time analytics: Contact Lens provides both real-time insights during the call, and post-call analytics after the conversation has ended and After Contact Work (ACW) is complete.

      If you choose this option, we recommend setting up alerts based on keywords and phrases that the customer may utter during the call. Contact Lens analyzes the conversation real-time to detect the specified keywords or phrases, and alerts supervisors. From there, supervisors can listen in on the live call and provide guidance to the agent to help them resolve the issue faster.

      For information about setting up alerts, see Alert supervisors in real-time based on keywords and phrases.

      If your instance was created before October 2018, additional configuration is needed to access real-time call analytics. For more information, see Service-linked role permissions for Amazon Connect.

  4. Choose the language. For a list of available languages for various Contact Lens features, see Supported languages.

    For instructions on using an attribute, see Use contact attributes.

  5. Optionally, enable redaction of sensitive data. For more information, see the next section, Enable redaction.

  6. Choose Save.

  7. If the contact is going to be transferred to another agent or queue, repeat these steps to add another Set recording and analytics behavior block with Enable Contact Lens for conversational analytics enabled.

Enable chat analytics

  1. In the Set recording and analytics behavior block, under Analytics, choose Enable Contact Lens conversational analytics, and Enable chat analytics.

    If you don't see this option, Contact Lens for Amazon Connect hasn't been enabled for your instance. To enable it, see Update instance settings.

  2. Choose the language. For a list of available languages for various Contact Lens features, see Supported languages.

    For instructions on using an attribute, see Use contact attributes.

  3. Optionally, enable redaction of sensitive data. For more information, see the next section, Enable redaction of sensitive data.

  4. Choose Save.

  5. If the contact is going to be transferred to another agent or queue, repeat these steps to add another Set recording and analytics behavior block with Enable Contact Lens for conversational analytics enabled.

Enable redaction of sensitive data

To enable redaction of sensitive data in a flow, choose Redact sensitive data. When redaction is enabled you can choose from the following options:

  • Redact all personally identifiable information (PII) data (all PII entities supported).

  • Choose which PII entities to redact from the list of supported entities.

If you accept the default settings, Contact Lens redacts all personally identifiable information (PII) it identifies, and replaces it with [PII] in the transcript. This option is shown in the following image:

Select PII entities to redact

Under the Data redaction section, you can select specific PII entities to redact, as shown in the following image:

Choose data redaction replacement

Under the Data redaction replacement section, you can choose the mask to be used as data redaction replacement, as shown in the following image:

For more information about using redaction, see Use sensitive data redaction.

Review sensitive data redaction for accuracy

The redaction feature is designed to identify and remove sensitive data. However, due to the predictive nature of machine learning, it may not identify and remove all instances of sensitive data in a transcript generated by Contact Lens. We recommend you review any redacted output to ensure it meets your needs.

Important

The redaction feature does not meet the requirements for de-identification under medical privacy laws like the U.S. Health Insurance Portability and Accountability Act of 1996 (HIPAA), so we recommend you continue to treat it as protected health information after redaction.

For the location of redacted files and examples, see Output file locations.

Dynamically enable Contact Lens using contact attributes

You can dynamically enable Contact Lens and the redaction of the output files based on the language of the customer. For example, for customers using en-US, you may want only a redacted file whereas for those using en-GB, you may want both the original and redacted output files.

  • Redaction: choose one of the following (they are case sensitive)

    • None

    • RedactedOnly

    • RedactedAndOriginal

  • Language: Choose from the list of available languages.

You can set these attributes in the following ways:

  • User defined: use a Set contact attributes block. For general instructions about using this block, see How to reference contact attributes. Define the Destination key and Value for redaction and language as needed.

    The following image shows how to use contact attributes for redaction. Note that Value is case sensitive.

    The following image show how to use contact attributes for language:

  • Use a Lambda function. This is similar to how you set up user-defined contact attributes. An AWS Lambda function can return the result as a key-value pair, depending on the language of the Lambda response. The following example shows a Lambda response in JSON:

    { 'redaction_option': 'RedactedOnly', 'language': 'en-US' }

What if the flow block fails to enable Contact Lens?

It's possible that the Set recording and analytics behavior block can fail to enable Contact Lens on a contact. If Contact Lens isn't enabled for a contact, check the flow logs for the error.

Multi-party calls and Contact Lens

We do not support multi-party calls in Contact Lens. For example, if there are more than two parties (agent and customer) on a call, the quality of the transcription and analytics, such as sentiment and categories, is degraded.