How do I send call recordings, transcript, and sentiment analysis from Amazon Connect to Salesforce?

0

Hi all, I've created an Amazon Connect instance and integrated with a Salesforce instance. I downloaded the Amazon connect lambda package and enabled call recording and transcripts and sentiment analysis. However, I am unable to create the contact trace record within Salesforce whenever a call is placed. Does anyone have any advice on how to get the contact trace record to create within Salesforce? Thank you!

2 Answers
0

Can you confirm you are using the Amazon Connect CTI adapter available on SF AppExchange. The Lambda package that you installed need to work with the CTI adapter.

If so, please follow the setup guide to setup how to stream Call Recording to SF. As well as transcript and sentiment from Contact Lens if they are enabled.

Contact Channel Analytics: https://amazon-connect.github.io/amazon-connect-salesforce-cti/docs/lightning/salesforce-lambdas/04-contact-channel-analytics

Contact Trace Record Import https://amazon-connect.github.io/amazon-connect-salesforce-cti/docs/lightning/salesforce-lambdas/05-contact-trace-record-import

Postcall Contact Lens Import https://amazon-connect.github.io/amazon-connect-salesforce-cti/docs/lightning/salesforce-lambdas/06-postcall-contact-lens-import

You might want to review the parameters you chose when installing the Lambda function https://amazon-connect.github.io/amazon-connect-salesforce-cti/docs/lightning/installation/04-salesforce-lambdas-manual-setup#install-the-amazon-connect-salesforce-lambda-package

Like CTRKinesisARN, ConnectRecordingS3BucketName, ContactLensImportEnabled, PostcallCTRImportEnabled, PostcallRecordingImportEnabled, etc

profile pictureAWS
answered 7 months ago
  • Hello, Yes I am using the adapter and I have followed these guides. I know that the lambda is working because I am able to create records from the contact flow in AWS by invoking the lambda function. However, the contact channel analytics record is still not getting created. Thank you, Venkat

  • To troubleshoot, you will need to check if your Kinesis Data Stream is getting any record, and then check your Lambda function call <name>-sfCTRTrigger-<id> which should have a trigger from Kinesis. And check Cloudwatch log to see if it is getting trigger. If so, follow the logic and see if you set the Contact Attribute in your flow correctly and the Lambda environment variable to true (https://amazon-connect.github.io/amazon-connect-salesforce-cti/docs/classic/cti-adapter/04-call-recording-playback/#cloudformation-template)

    If some of the parameters is not right and you adjusted afterwards, I found it much easier to re-deploy and pick the right parameter rather then figure out ALL the lambda functions and which one to enable

-2

Hello Ven,

To send call recordings, transcripts, and sentiment analysis data from Amazon Connect to Salesforce, you can set up an integration between the two platforms using various AWS services like Lambda, Amazon S3, and AWS Step Functions. Here's a step-by-step guide to help you achieve this:

Step 1: Set Up Amazon S3 Bucket

  1. Create an Amazon S3 bucket where you will store call recordings, transcripts, and sentiment analysis results. Make sure to configure the bucket's permissions to allow access from your AWS services.

Step 2: Enable Recording in Amazon Connect

  1. In your Amazon Connect instance, enable call recording, transcripts, and sentiment analysis for your contact flows.

Step 3: Create a Lambda Function

  1. Create an AWS Lambda function that will be triggered by Amazon Connect to process call records. You can use the Amazon Connect Lambda blueprint or write a custom function to suit your needs.

  2. In the Lambda function, use the AWS SDK to interact with Amazon S3 to upload the call recording, transcript, and sentiment analysis results to the S3 bucket created in Step 1.

  3. The Lambda function can also be responsible for creating a Salesforce contact trace record using the Salesforce API.

Step 4: Set Up Amazon Connect Contact Flow

  1. In your Amazon Connect contact flow, configure the contact flow to invoke the Lambda function created in Step 3 when a call is completed or when you want to process the data.

  2. You can use the "Invoke AWS Lambda function" block in your contact flow to trigger the Lambda function.

Step 5: Create Salesforce Contact Trace Record

  1. In your Lambda function, use the Salesforce API to create a contact trace record within Salesforce. You will need to authenticate your Lambda function with Salesforce by setting up the necessary OAuth authentication.

  2. Map the relevant call data, transcripts, and sentiment analysis results as fields in the Salesforce contact trace record.

Step 6: Test the Integration

  1. Test your integration by making calls through Amazon Connect and ensuring that call recordings, transcripts, and sentiment analysis data are being processed and sent to Salesforce.

Step 7: Monitor and Debug

  1. Implement logging and monitoring in your Lambda function and contact flow to troubleshoot and debug any issues that may arise during the integration.

Step 8: Secure Data Handling

  1. Ensure that you handle sensitive data (such as recordings) in a secure manner and comply with any regulatory requirements (e.g., GDPR) for data privacy.

Step 9: Error Handling and Retry Mechanism

  1. Implement error handling and retry mechanisms in your Lambda function to handle scenarios where data cannot be uploaded to S3 or where Salesforce API requests fail.

By following these steps and setting up the integration carefully, you should be able to send call recordings, transcripts, and sentiment analysis data from Amazon Connect to Salesforce, and create contact trace records within Salesforce for each call.

Please give a thumbs up if it helps

profile picture
answered 7 months ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions