Increasing agent productivity with generative AI in Amazon Connect

Introduction

Contact center agents face a variety of challenges when handling customer interactions. Whether it’s troubleshooting a technical issue, resolving a billing dispute, or simply providing helpful, accurate information, agents need to be effective across a wide range of topics. This often requires months, or even years, of experience becoming familiar with the various issues and technologies needed to assist customers.

However, the work doesn’t end when the interaction is over. Agents also need to complete critical after-call work like updating customer records, logging case notes, and following up on any outstanding issues. This administrative work is time consuming but essential for maintaining high customer satisfaction (CSAT) scores and operational efficiency.

With an ever-increasing number of customer inquiries across multiple channels, agents are under constant pressure to handle a high volume of interactions quickly while still providing an excellent customer experience. This mounting workload can lead to burnout, higher turnover rates, and inconsistent service levels.

Generative artificial intelligence (AI) technologies present an opportunity to augment human agents and drive substantial productivity gains throughout the customer service lifecycle. By automating mundane post-contact tasks and providing agents real-time information assistance, generative AI acts as a force multiplier – empowering agents to work faster and smarter. In this blog post, we’ll explore how generative AI-powered capabilities in Amazon Connect can boost agent productivity during and after customer interactions.

Let’s start by taking a look at how Amazon Q in Connect can assist agents during contacts by providing real-time assistance and recommending relevant guided workflows. Afterwards, we’ll review how Amazon Connect Contact Lens reduces the manual effort required by agents to complete after-contact work.

Amazon Q in Connect real-time agent assistance

Amazon Q in Connect uses generative AI to deliver agents suggested responses and actions to address customer questions, providing faster issue resolution and improved customer satisfaction. Using conversational analytics and natural language processing (NLP), Amazon Q in Connect can automatically detect customer issues and respond to conversational search when an agent needs additional information.

Amazon Q in Connect now not only provides agents solutions derived from knowledge articles, wikis, and FAQs but also recommends related step-by-step guides to provide the appropriate steps needed to complete the task at hand. Step-by-step guides in Amazon Connect help contact center agents quickly and efficiently resolve customer inquiries. These customizable guides provide a tailored workflow, walking agents through the necessary steps to address customer issues. With no-code, drag-and-drop workflows, administrators can design guides for various customer interactions, such as processing orders, managing payments, or handling returns. These guides dynamically adapt based on the customer’s context, displaying relevant information about the issue and guiding the agent on the appropriate actions to take. Step-by-step guides are already available in the Amazon Connect agent workspace. Amazon Q in Connect surfaces this guided assistance, recommending the guides alongside generated responses and solutions. This helps decrease training time for new agents and boost productivity for experienced ones. Agents always know the next best step, leading to faster, more accurate resolutions.

Associating step-by-step guides with knowledge articles allows you to surface relevant information, enabling agents to answer questions and take the next steps needed to swiftly resolve the issue, without leaving their agent workspace. By having contextual information and best practice guides at their fingertips, agents can work through issues faster while delivering more consistent and accurate resolutions. This improves critical metrics like average handle time, first contact resolution rates, and customer satisfaction scores.

Let’s look at an example of how Amazon Q in Connect enhances agent productivity:

Example: Amazon Q in Connect at work in a customer interaction

An agent is speaking with a customer who needs to update their payment method. After a customer requests ‘update payment method,’ Amazon Q in Connect assists agents in real-time by recommending a response and an associated guide. The guide streamlines the process, walking agents through precise steps to update the customer’s billing information within their workspace, eliminating the need to juggle multiple tabs and applications.

The guide is delivered as an interactive flow alongside the agent’s other applications. This allows them to seamlessly follow along without losing context or disrupting their workflow.

Amazon Q in Connect can also help an agent handle an issue as the conversation changes topic. For example, if a customer calls about making a return, Amazon Q in Connect might initially recommend the general “Returns and Exchanges” guide. But once the agent identifies it’s a defective product issue, Amazon Q would then display the specific “Defective Product Return” workflow. No more wasted time searching through different articles – the agent has the exact instructions they need front and center.

After a customer interaction, an agent needs to submit a refund request. Agents can chat with Amazon Q in Connect, saying ‘I need to process a refund.’ It surfaces relevant knowledge base information and the approved process flow. Amazon Q processes natural language requests, returning accurate information and the steps to complete the task.

By having relevant knowledge on tap, Amazon Q in Connect allows agents to rapidly find information, follow procedures, and complete tasks with minimal effort or application switching. This eliminates one of the biggest productivity drains in the contact center. Beyond boosting performance metrics, Amazon Q in Connect reduces training time for new agents and provides a more intuitive desktop experience. This increases employee satisfaction and reduces costly turnover. With generative AI-powered features like Amazon Q in Connect, contact centers can finally fully leverage their knowledge bases and empower agents to deliver faster, smarter customer service.

How to associate step-by-step guides with knowledge content

In order for Amazon Q in Connect to recommend step-by-step guides alongside recommendations, you must first associate a guide with the article. An article can have a single guide associated with it, but a guide can be associated with multiple articles, if desired. Administrators can associate relevant guides with their knowledge content using the available CreateContentAssociation Amazon Q in Connect API. This API allows you to enhance knowledge articles with guided workflows, forms, and embedded third-party applications needed to effectively resolve the customer task. Combined with the other Amazon Q in Connect APIs (GetContentAssociation, ListContentAssociation, DeleteContentAssociation) you have full control over when, where, and how to extend the assistance provided by Amazon Q in Connect into actionable steps for agents.

Amazon Connect Contact Lens for automating post-contact work

It’s not just during a contact where generative AI can assist agents. Amazon Connect Contact Lens provides conversational analytics and quality management capabilities that enables the contact center to monitor, measure, and continuously improve contact quality and agent performance for a better overall customer experience.

Contact Lens provides generative AI-powered post-contact summaries, which summarize long customer conversations into succinct, coherent, and context rich contact summaries. This helps supervisors improve the customer experience by getting faster insights when reviewing contacts, saving time on quality and compliance reviews, and more quickly identifying opportunities to improve agent performance. Today, these summaries are also available to help agents perform their after-contact work (ACW) more effectively. With ACW summaries, agents no longer need to take manual notes for every customer conversation; a process that could be both error-prone and time consuming. This capability harnesses the power of generative AI to automatically generate context-rich contact summaries immediately after a conversation ends. Within seconds, the full interaction is analyzed and a detailed summary capturing key discussion points, issues raised, actions taken, and other critical context from the contact is presented to the agent and the supervisor. The generative AI-powered summaries provide a complete record of the contact that can be seamlessly attached to the customer record, eliminating this tedious step from the agent workflow. This reduces after-contact work and allows agents to maximize their time focusing on providing stellar customer service.

Amazon Connect Contact Lens offers the flexibility to access chat and voice summaries via API, Kinesis Streams, Connect Control Panel (CCP), and the Contact Details page. This enables seamless integration with other applications, such as Amazon Connect Cases or Salesforce, allowing agents to quickly update customer records and ensure data consistency across platforms. Let’s discuss the different ways to take advantage of these ACW summaries.

Example: Amazon Connect Contact Lens ACW summaries streamlines post-contact work

After an agent completes a customer interaction, they are tasked with completing their wrap-up work before moving onto the next customer conversation. To help support this, Amazon Connect Contact Lens provides a summary of the conversation in the CCP. Currently supported for voice contacts, the summary is available in the CCP seconds after the contact disconnects and the agent is moved into ACW. This view provides the full transcript and key highlights, but includes the conversation summary for easy reference at the top. The agent can refer to this summary for completing ACW activities, or copy it directly into wrap-up fields as needed.

For supervisors, having visibility into what is going on throughout the contact center is critical. Fortunately, the contact details page provides the ACW summary as soon as it is available – even while the agent is still in ACW before they close the contact. Supported across both voice and chat channels, it allows supervisors, agents with permission, or other Amazon Connect users, to quickly see a summary of the contact.

There may be other applications of this summary within the contact center. The Contact Lens ListRealTimeContactAnalysisSegments (voice) and ListRealTimeContactAnalysisSegmentsV2 (chat) APIs allow you to return the generative AI-powered conversation summary seconds after the contact has ended. These APIs can be integrated into agent workflows, for example to include the summary in step-by-step guides to reference while completing ACW activities. The API can also be used to accomplish ACW tasks in other applications used by the agent.

In addition to the methods described previously, these summaries can be used to increase productivity in other applications used across the contact center. Supported across both voice and chat channels, by streaming contact summaries directly into Amazon Kinesis Data Streams, contact centers can build analytics tools and agent experience enhancements based on continuous flow of Contact Lens data, particularly summary content, for all enabled contacts. By directly integrating Amazon Kinesis with services like AWS Lambda and Amazon Data Firehose, customers can leverage this data to address business challenges, seamlessly integrating with applications and systems like Salesforce customer relationship management (CRM). This integration allows for automated updates, ensuring that customer data remains up-to-date and accessible across various touch-points without manual agent action required, enhancing overall operational efficiency and customer satisfaction.

Demo

Want to see how Amazon Q in Connect and Contact Lens help an agent during an interaction? Take a look at the following demo:

Conclusion

While contact center agents often face immense workloads and administrative burdens that can hinder their productivity and lead to burnout, generative AI capabilities provide a powerful solution to this challenge.

By automating tedious tasks, providing real-time assistance during interactions, and generating comprehensive post-contact summaries, generative AI-powered Amazon Connect empowers agents to work more efficiently, deliver faster resolutions, and improve overall customer satisfaction. With contextual knowledge at their fingertips and streamlined workflows, agents can focus on what matters most – delivering personalized, high-quality service.

As businesses continue to prioritize customer experience, investing in generative AI technologies like Amazon Q in Connect and Amazon Connect Contact Lens will be crucial for driving productivity gains, reducing operational costs, and fostering a more engaged and effective workforce. By harnessing the power of AI, contact centers can unlock new levels of efficiency and deliver exceptional customer experiences at scale.

Resources to get started with Amazon Q in Connect

• Do you want to learn how to activate Amazon Q in Connect in your instance? The Amazon Connect Admin Guide provides a comprehensive overview of enabling Q in Connect and its features. Learn more at Enable Amazon Q in Connect for your instance.
• Do you want to get hands-on with Amazon Q in Connect? Follow this Amazon Q in Connect workshop to learn how to activate, set up your domain, connect your content, and deliver those AI-generated response and solutions to your agents.
• Do you want to integrate existing step-by-step guides with Amazon Q in Connect content? Amazon Q in Connect supports association of views to your content. Learn more at Integrate Amazon Q in Connect with step-by-step guides.

Resources to get started with Amazon Connect Step-by-step Guides

• Do you want to get started building your first step-by-step guide in Amazon Connect? Follow this step-by-step guides workshop to learn more about how to build, deploy, and test a sample guides that interacts with Amazon Connect attributes to provide a personalized, dynamic, and contextual experience.
• Do you want to dive-deep with step-by-step guides? Learn more in the Amazon Connect Administrator Guide on how to onboard your third-party applications.

Resources to get started with Amazon Connect Contact Lens

• Do you want to get hands-on with Amazon Connect Contact Lens?
  Follow the Amazon Connect Contact Lens workshop to learn how to setup and explore the capabilities available within Amazon Connect Contact Lens.
• Do you want to learn more about generative AI-powered post-contact summaries? Learn more in the Amazon Connect Administrator Guide on how to enable, view, and use these summaries in your contact center.
• Do you want to dive-deep with step-by-step guides? Learn more in the Amazon Connect Administrator Guide on how to onboard your third-party applications.

Ready to transform your customer service experience with Amazon Connect? Contact us.

Author bio

Alex Schrameyer (he/him) is a Worldwide Solutions Architect Lead for Agent Experience at Amazon Web Services (AWS) based in Indianapolis, Indiana. He believes that exceptional agent experiences are the cornerstone of outstanding customer service, and focuses on creating environments where agents excel and customers are delighted. Alex enjoys traveling around the world, and you might find him at your local baseball stadium or theme park.

TP Kohli is a Senior Product Manager focused on building, managing, and scaling AI/ML applications. He is currently responsible for delivering the best conversational analytics and quality management capabilities on Amazon Connect for contact center customers using Generative AI. TP loves solving customer use cases, earn trust with customers, and deliver the best user experience that help contact center customers monitor, measure, and continuously improve contact quality and agent performance for a better overall end-customer experience.