Amazon Lex Integration

Collect Amazon Lex conversation events and logs to analyse bot performance, intent recognition, and customer self-service interactions.

What is Operata Amazon Lex Integration?

Operata Amazon Lex Integration collects conversation events and logs from Amazon Lex V2 bots used within your Amazon Connect contact flows. This data is ingested into the Operata platform and correlated with Contact Trace Records, IVR flow logs, and agent experience metrics.

By bringing Lex data into Operata, you gain visibility into what customers say to your bots, how well intents are recognised, and where conversations break down before reaching an agent.

Why Collect Lex Data?

Bot interactions are a critical part of the customer journey, but they are often a blind spot in contact centre analytics. Lex Integration enables you to:

  • Measure bot effectiveness - track intent recognition rates, fallback frequency, and successful self-service completions
  • Identify misunderstood intents - surface utterances that are being misclassified or falling through to the default intent
  • Correlate bot interactions with agent outcomes - understand how bot conversations influence downstream queue selection, handle time, and customer satisfaction
  • Monitor API performance - detect latency or failures in Lambda fulfillment functions and external API calls triggered by Lex

Prerequisites

Before deploying the Amazon Lex Integration stack, ensure you have:

  • An active Operata subscription
  • An AWS account with appropriate IAM permissions to deploy CloudFormation stacks
  • Amazon Lex V2 bots integrated with your Amazon Connect contact flows
  • Lex conversation logging enabled for the bots you want to monitor

Tip: Conversation logging can be configured per bot alias in the Amazon Lex V2 console. Both text and audio logs can be enabled, though Operata primarily consumes text-based conversation log data.

Architecture Overview

Amazon Lex Integration uses CloudWatch Logs and EventBridge to capture Lex conversation events and stream them to the Operata platform.

Data flow:

  1. Amazon Lex writes conversation logs to CloudWatch when conversation logging is enabled
  2. Lex API events are captured via CloudTrail and EventBridge
  3. The Operata CloudFormation stack creates pipelines that subscribe to these event sources
  4. Events are streamed to the Operata API endpoint via EventBridge API Destinations
  5. Operata correlates Lex data with CTR, IVR flow logs, and agent experience data

AWS Services Deployed

The CloudFormation template provisions the following resources in your AWS account:

ServiceQuantityPurpose
EventBridge Pipe1Streams Lex events to Operata
API Destination1Operata platform endpoint
IAM Role1Service permissions for the pipeline
IAM Policies2Access control for event sources and API Destination
Secrets Manager Secret1Stores the Operata API key

Deployment

Preparation

Provide Operata with the following details to generate your custom CloudFormation template:

  • Region of your AWS Account
  • CloudWatch Log Group name for Lex conversation logs (conversation logging must be enabled - see Prerequisites)
    • usually follows the format of /aws/lex/<log group name>
    • If you have multiple lex bots and want to setup for a specific bot, then go to Amazon Lex > <select the bot> from the list > (right hand panel) > Deployment > Aliases > (copy) <CloudWatch Log Group Name> value
  • Operata Group ID
    • we can assist on this if you are unsure

Installation

  1. Operata will provide a CloudFormation quick-create link pre-filled with your configuration.
  2. Click the link to open the CloudFormation console.
  3. Enter your Operata API Key in the parameter field. (How to create an API Key)
  4. Review the stack, agree to the terms, and deploy.
  5. Wait for the stack to reach CREATE_COMPLETE status (typically under 5 minutes).

Post-Deployment Testing

  1. Make a test call that triggers your Lex bot through an Amazon Connect contact flow.
  2. Interact with the bot (speak or enter DTMF inputs depending on your flow).
  3. Log into the Operata console.
  4. Locate your call and verify that Lex conversation data appears alongside the Contact Trace Record and IVR flow logs.

Note: Lex events may take up to 5 minutes to appear in the Operata console after a call completes.

Data Included

The integration captures Lex conversation events including:

  • Session ID and bot identifiers
  • Intent names and recognition confidence scores
  • Slot values (subject to privacy controls)
  • Utterance text (subject to privacy controls)
  • Session attributes
  • Fulfillment status and Lambda response data
  • Conversation timestamps

For the full Lex logging and monitoring data model, see the AWS documentation.

Privacy Considerations

Lex conversation data may contain customer PII, particularly in utterance text and slot values. For example, a customer might say their account number or address during a bot interaction.

Operata supports privacy controls for Lex data through the same allow-list and deny-list configuration used for CTR redaction. Contact Operata support to discuss the appropriate redaction configuration for your Lex data.

For details on configuring privacy controls, see API Data Privacy Controls.

Charges

All AWS service charges for the resources deployed by this stack are the responsibility of the customer. Amazon Lex Integration is included in all Operata plans at no additional Operata cost.

Primary cost factors:

  • EventBridge event processing
  • CloudWatch Logs data transfer
  • Secrets Manager secret storage

Data Transfer: You may incur additional data transfer charges for data out to the internet. See AWS Data Transfer pricing for details.

Deploying Alongside Other Stacks

Amazon Lex Integration is designed as an independent CloudFormation stack. You can deploy it alongside the existing CTR Collection, Contact Lens, and IVR Flow Log Collection stacks. Each stack operates independently and can be added or removed without affecting the others.

Support

ChannelDetails
ChatIn-app
Email[email protected]
PhoneOn Application