Contact Lens for Amazon Connect

Real-time conversational analytics and quality management powered by machine learning

Contact Lens for Amazon Connect, a feature of Amazon Connect, provides a set of conversational analytics and quality management capabilities, powered by machine learning, that helps understand and classify the sentiment, trends, and compliance of your conversations. These contact center analytics can be used to take action, addressing customer issues, improving agent performance, and identifying crucial company and product feedback. You can more easily search call and chat transcripts, analyze sentiment, identify issues, and monitor agent performance in real-time. Contact Lens also helps you define and assess agent performance criteria (such as adherence to required scripts) and automatically complete evaluation forms. This reduces the need for managers to do manual reviews while more easily identifying coaching needs for your agents.

Getting Started with Contact Lens for Amazon Connect (6:24)

Understand sentiment, conversation characteristics, and agent compliance risks surfaced using natural language processing (NLP) on transcripts generated by ML-powered conversational analytics for calls and chats. This can verify that standard greetings and sign-offs are used, help train agents, and replicate successful interactions. Set real-time alerts to flag agent coaching opportunities and discover customer insights with detailed analysis in the analytics dashboard. Use real-time data streams to build customized dashboards that include sentence-by-sentence transcripts, sentiment analysis, and categories from customer conversations.

Improve agent productivity and customer service

Save your agents’ valuable time on every customer interaction by summarizing key parts of the conversation, reducing the need for them to take detailed notes and instead focus on solving the customer’s issue. Managers can view contact summaries alongside the contact details to quickly understand the context of an interaction, help address follow-up tasks, and provide precise feedback to agents.

Automate follow-ups to improve customer experience

You can use the automated contact categorization to create Amazon Connect Tasks for customer requests. Automating tasks helps to address customer follow-up promises, such as scheduling a call back or initiating a refund. This helps you deliver high-quality customer service while improving agent productivity.

Enhance contact center security and compliance

Detect and redact sensitive customer data such as credit card details, addresses, and social security numbers from audio recordings and transcripts. You can also improve agent compliance with company policies or regulatory requirements by tracking all customer conversations for script adherence using categorization-based on criteria you determine. For example, you can track words or phrases used in required disclaimers, greetings and sign-offs.

Use evaluation forms to improve agent performance

Using evaluation forms (preview), you can review conversations alongside contact details, recordings, transcripts, and summaries, without the need to switch applications. Conversational analytics automatically populates evaluation results scoring criteria like script adherence, sensitive data collection, and customer greetings to reduce the time you spend identifying and coaching agents, helping them to perform at their best.

Contact Lens customers

Intuit logo
Origin logo
Accolade logo
University of Auckland
Energy Ogre logo
News Corp logo
Amaysim logo
See customer stories »


Automated contact categorization

Track customer conversations for compliance with company policies or regulatory requirements. Define and manage categories directly based on your specific criteria (such as words, phrases, conversational characteristics like sentiment, interruptions, and non-talk time) within Amazon Connect using rules. Rules are an ML-based contact categorization engine that can automatically label contacts for real-time and post-contact scenarios.

Automated evaluation (preview)

Get evaluation results automatically populated into your forms based on conversational analytic outputs such as sentiment or transcripts. This helps you save time from manually looking for hot spots in the recordings or transcripts.

Contact search

Contact search helps you quickly search a few relevant contacts from many contacts across the contact center. You can use filters such as agent name, queue name, ML-enabled conversational analytics (such as specific keywords, categories, and sentiment score), contact attribute, and many more to drill down to those relevant contacts in less time. This helps you understand customer trends and insights and how to improve customer satisfaction.

Call recording

Access live and recorded conversations for your contact center calls to address use cases such as monitoring agent compliance, evaluating contact quality, and identifying calls for training purposes. These call recordings are stored in an Amazon S3 bucket, which you can consume outside of Amazon Connect. They are also visualized within Amazon Connect on the contact details page.

Contact summarization

Automatically identify key parts of the customer conversation, assign tags (such as issue, outcome, or action item), and display a summary that can be expanded to view the full transcript of the contact.

Contact details and analytics

View all conversational analytics at an individual contact level, including transcripts, agent and customer sentiment, contact categories, non-talk time, response time, and other conversation characteristics, to detect issues and customer trends. Contact details also helps you access details such as contact id, agent name, disconnect reason, and start and end time.

Custom vocabularies

Improve the accuracy of speech recognition for terminology (such as product names and brand names) by expanding the vocabulary of Contact Lens’ speech-to-text engine. You can provide a list consisting of domain-specific words and phrases, words that aren’t being recognized correctly, and proper nouns. 

Sentiment analysis

Capture and analyze the sentiment of words being spoken by the customer through ML-powered natural language processing (NLP). It will generate a score between -5 (most negative) to +5 (most positive). 

Email notification 

Receive real-time email notifications when configurable conditions of a rule (such as customer sentiment) get initiated. This will help you identify and intervene on contacts where agents might need additional support and provide guidance to deliver better end-customer experiences.

Evaluation forms and contact scoring (preview)

Define and create a set of agent performance evaluation forms and complete the evaluations side-by-side with call recordings, transcripts, and conversational analytics outputs such as contact categories, sentiment scores, and issues detected. Get a contact scored based on the evaluation result instantly completed for your review. 

Sensitive data redaction

Remove sensitive data (such as names, addresses, credit card details, and social security numbers) from both the call or chat transcripts and audio recordings.

Real-time alerts 

Create rules to flag any customer experience issue in real time, with categories based on keywords, sentiment, and phrase matching. This automatically alerts your supervisors in real time when they need to assist an agent on live contacts so they can provide guidance through chat or have the agent transfer the call.

Real-time data streams Access real-time analytics using data streams to provide issue detection, sentence-by-sentence transcripts, sentiment analysis, and categories for ongoing customer conversations with low latency.


Q: How can I learn more about Amazon Connect?

For more information, see Amazon Connect.

Q: How much does Contact Lens for Amazon Connect cost?

For pricing information, see Amazon Connect Pricing

Q: How do I get started?

To get started with Contact Lens for Amazon Connect, see Enable Contact Lens for Amazon Connect. The documentation provides instructions for how to turn on Contact Lens within your Amazon Connect instance.

Q: How do I access data in Contact Lens for Amazon Connect for use outside of Amazon Connect?

The metadata generated by Contact Lens for Amazon Connect (including call transcript, sentiment analysis, non-talk time, categorization labels, talk speed, and interruptions), along with the call recordings for each contact, will be accessible in your Amazon Simple Storage Service (S3) bucket. This data will be linked to Contact Trace Records (CTR) and can be used in BI tools like Amazon QuickSight and Tableau. You can create custom visualizations that fuse the CTR data with data from other systems (such as CRM). Finally, your analytics teams can also use this data to create custom machine learning (ML) models with Amazon SageMaker.

Q: How does Contact Lens for Amazon Connect relate to Amazon Transcribe and Amazon Comprehend?

Contact Lens is an out of the box feature for Amazon Connect that leverages Amazon Transcribe to generate call transcripts and Amazon Comprehend to apply natural language processing (NLP) on these transcripts, with no coding required. This approach helps organizations evaluate their customer experience using Contact Lens for Amazon Connect, without requiring expertise in Amazon Transcribe or Amazon Comprehend.

Q: What languages does Contact Lens for Amazon Connect support?

To view a list of languages Contact Lens for Amazon Connect currently supports for post-call analytics, post-call redaction, real-time analytics, and real-time redaction, see Languages supported by Amazon Connect. We will continue to add support for more languages.

Q: What should I know before using sensitive data redaction?

The redaction feature is designed to identify and remove sensitive data. However, due to the predictive nature of ML, it might not identify and remove all instances of sensitive data in a transcript generated by Contact Lens for Amazon Connect. We recommend reviewing the results for accuracy after enabling sensitive data redaction to verify they meet your needs.

Q: Can sensitive data redaction be used for healthcare data or protected health information?

The redaction feature is not intended to be used to de-identify healthcare data or to remove references to protected health information.

Q: Can I programmatically access the real-time capabilities of Contact Lens for Amazon Connect?

Yes. The real-time capabilities of Contact Lens are available through either the Amazon Connect user interface or a synchronous real-time API that help you to build customized solutions for use cases like agent transfers.

Q: Is the content processed by Contact Lens for Amazon Connect moved outside the AWS Region where I am using Contact Lens for Amazon Connect?

Any content processed by Contact Lens for Amazon Connect is encrypted and stored at rest in the AWS Region where you are using Contact Lens for Amazon Connect. Unless you opt out as provided below, some portion of content processed by Contact Lens for Amazon Connect might be stored in another AWS Region. This would solely be with the continual improvement and development of your Contact Lens for Amazon Connect experience and other Amazon machine-learning/artificial-intelligence technologies. You can request deletion of content associated with your account by contacting AWS Support. Your trust, privacy, and the security of your content are our highest priority. We implement appropriate and sophisticated technical and physical controls, including encryption at rest and in transit, designed to prevent unauthorized access to, or disclosure of, your content and verify that our use complies with our commitments to you. See Data Privacy FAQ for more information. Your content will not be stored in another AWS Region if you opt out of having your content used to improve and develop the quality of Contact Lens for Amazon Connect and other ML and artificial intelligence (AI) services on AWS. You can opt out of having your content used to improve and develop the quality of Contact Lens for Amazon Connect and other Amazon machine-learning/artificial-intelligence technologies by using an AWS Organizations opt-out policy. For information about how to opt out, see AI services opt-out policy.

Q: Once I have Contact Lens enabled, how do I get started with the evaluation forms?

After you have set up Contact Lens you can configure your instance for agent performance evaluation forms by configuring your security profiles, which set permissions for specific users to define, create, or review contact evaluations. Refer to the Administration Guide to learn more.

Q: What customer contacts can be reviewed with the evaluation forms? Are conversational analytics required to evaluate the agent performance of a contact?

You can use the evaluation forms for all contacts within your Amazon Connect instance, including conversations over voice calls, chats, and tasks. You can manually evaluate contacts without conversational analytics, or enable conversational analytics to automate the evaluation process and get results pre-populated within the form. We do not support evaluating customer interactions outside of your Amazon Connect instance.

Regions available:

Conversational analytics (post-contact) for speech and chat and evaluation forms are available in the US West (Oregon), US East (N. Virginia), Canada (Central), Europe (London), Europe (Frankfurt), Asia Pacific (Singapore), Asia Pacific (Seoul), Asia Pacific (Tokyo), and Asia Pacific (Sydney) Regions. Conversational analytics (real-time) for speech are available in the US West (Oregon), US East (N. Virginia), Canada (Central), Europe (London), Europe (Frankfurt), Asia Pacific (Seoul), Asia Pacific (Tokyo), and Asia Pacific (Sydney) Regions.

Learn more about Amazon Connect
Learn more about Amazon Connect

Visit the product overview page.

Learn more 
Sign up for an AWS account
Sign up for a free account

Instantly get access to the AWS Free Tier. 

Sign up 
Start building on the console
Start building on the console

Get started building with Contact Lens for Amazon Connect in the AWS Management Console.