Home AI News Transforming Customer Experience with AI for Call Analytics at AWS re:Invent 2023

Transforming Customer Experience with AI for Call Analytics at AWS re:Invent 2023

Introduction
Key Contact Center Challenges
Personas and Their Roles
- Customers
- Agents
- Managers and Supervisors
Solutions for Contact Center Challenges
- Self-Service Virtual Agents
- Real-Time Call Analytics and Agent Assist
- Conversation Analytics
Benefits for Customers and Businesses
- Case Study: W Bank
- Case Study: Magellan Health
- Case Study: State Auto Insurance
- Case Study: TSB Bank
Amazon Connect and AWS CC I Solutions
Overview of Post Call Analytics
New Features of Amazon Transcribe
Principal Financial Group's Journey with PC A
Approach and Deployment Phases
- Technical Deployment
- Topic Hierarchy Definition
- Reporting Enhancements and Additions
- Virtual Assistants and Emerging Theme Detection
Improving Customer Experience with PC A
- Holistic View of Customer Interactions
- The Power of Topics and Intents
- Omni-Channel Experience
- Roadmap for Future Enhancements
- Intelligent Agents and Q&A Bot
Conclusion
Resources and Further Information

Introduction

In this article, we will be discussing the impact of Generative AI in the contact center space and how it can be used to enhance customer experience. We will explore the key challenges faced by contact centers, the different personas involved, and the solutions that can alleviate these challenges. Additionally, we will examine the benefits of using generative AI through real-life case studies. We will also provide an overview of Amazon Connect and AWS CC I Solutions for contact centers. Furthermore, we will dive into the concept of post-call analytics and explore the new features of Amazon Transcribe. To illustrate the practical application of generative AI in contact centers, we will take a look at the journey of Principal Financial Group and how they have leveraged PC A for improved customer experience. Lastly, we will discuss the roadmap for future enhancements and the potential of intelligent agents and Q&A bots in contact centers.

Key Contact Center Challenges

Contact centers face several challenges in providing optimal customer service. One of the primary challenges is meeting customer expectations. With the rise of self-service solutions, customers now prefer quick and efficient resolutions to their queries. Agents, who are the face of the company's customer service department, often find themselves overburdened with a high volume of calls and administrative tasks. This affects the quality of customer interactions and increases call resolution time. Managers and supervisors also struggle with analyzing the vast amount of data collected from customer conversations to derive actionable insights. In the following sections, we will explore solutions and strategies to overcome these challenges and improve overall contact center performance.

Personas and Their Roles

To better understand the contact center ecosystem, it is essential to identify the key personas involved and their day-to-day roles. The three primary personas in contact centers are customers, agents, and managers/supervisors.

Customers

Customers are the most crucial persona in the contact center space. They are the ones seeking assistance and solutions to their queries or problems. Today, customers expect quick and convenient self-service solutions. They prefer chatbots or conversational assistants that can provide immediate resolutions. Therefore, it is essential for businesses to adopt self-service virtual agents to meet customer expectations and enhance their overall experience.

Agents

Agents play a critical role in the contact center as they are responsible for delivering customer service. However, agents often face challenges due to the high volume of calls and administrative tasks. This results in decreased productivity and increased call resolution time. To better support agents and improve their efficiency, real-time call analytics and agent assist solutions can be implemented. These solutions provide agents with Prompts and insights during ongoing conversations, helping them find answers faster and deliver better customer service.

Managers and Supervisors

Managers and supervisors are responsible for overseeing the contact center operations and ensuring optimal performance. They Collect vast amounts of data on a daily basis but often struggle to analyze and derive actionable insights from this data. To address this challenge, conversation analytics solutions can be employed. These solutions analyze all customer conversations and provide insights on overall call sentiment, agent performance, emerging business trends, and customer feedback. With these insights, managers and supervisors can make informed decisions to improve customer experience and boost business performance.

Solutions for Contact Center Challenges

To address the key challenges faced by contact centers, several solutions leveraging generative AI have emerged. These solutions focus on improving customer experience, enhancing agent productivity, and enabling managers to extract valuable insights from customer conversations. Let's explore some of these solutions in Detail.

Self-Service Virtual Agents

Self-service virtual agents, such as conversational IVRs and chatbots, powered by generative AI, can provide customers with quick and convenient solutions. These virtual agents are equipped with extensive knowledge bases similar to those used by company agents. Customers can find answers to their queries at any time, improving their overall experience and reducing the need for human assistance.

Real-Time Call Analytics and Agent Assist

Real-time call analytics and agent assist solutions offer valuable insights during ongoing customer conversations. These solutions use generative AI to analyze the sentiment and intent of customers, providing agents with prompts and suggestions in real-time. This empowers agents to respond faster and more accurately, resulting in improved efficiency, reduced call resolution time, and enhanced customer satisfaction.

Conversation Analytics

Conversation analytics solutions help managers and supervisors derive insights from customer conversations. These solutions analyze millions of calls and provide comprehensive data on call sentiment, agent performance, emerging business trends, and customer feedback. By analyzing all customer interactions, businesses can identify areas for improvement, make data-driven decisions, and enhance overall performance.

Benefits for Customers and Businesses

Several organizations have already witnessed significant benefits by implementing generative AI solutions in their contact centers. Let's take a look at some real-life case studies to understand the impact of these solutions on customer experience and business performance.

Case Study: W Bank

W Bank implemented a self-service conversational AI platform provided by AWS, which resulted in a 90% reduction in customer call duration for simple inquiries, such as balance inquiries. The average call duration decreased from 4.5 minutes to just 28 seconds. Additionally, the bank successfully contained 30% of their calls through self-service solutions, freeing up agent resources for more complex customer issues.

Case Study: Magellan Health

Magellan Health utilized real-time call analytics and agent assist solutions, reducing agent training time by 3 to 5 days. This may seem like a small improvement, but with over 2.2 million calls per year, the company saved around 4,400 hours of agent training time, resulting in significant cost savings.

Case Study: State Auto Insurance

State Auto Insurance implemented a post-call analytic solution, enabling them to analyze all of their customer calls. This comprehensive analysis helped them identify call intents and improve customer experience by directing calls to the right agents with the appropriate expertise. As a result, the company saved approximately $800,000 in operational expenses.

Case Study: TSB Bank

TSB Bank transitioned from analyzing only 10% to 100% of their calls using a post-call analytic solution. This allowed them to identify over 800 call intents, which helped improve customer experience by efficiently routing calls to the right agents Based on call intent. By focusing on customers' needs, TSB Bank enhanced overall satisfaction and loyalty.

These case studies demonstrate the significant benefits that generative AI solutions can bring to contact centers. From reducing call duration to improving agent efficiency, implementing these solutions can result in substantial cost savings, enhanced customer satisfaction, and improved business performance.

Amazon Connect and AWS CC I Solutions

For organizations looking to leverage generative AI in their contact centers, Amazon Connect and AWS CC I Solutions offer flexible options to get started quickly. Amazon Connect is a contact center solution that allows businesses of all sizes to provide superior customer experience. It offers multiple features and services, including automatic speech recognition provided by Amazon Transcribe and generative AI capabilities provided by Amazon Bedrock.

For those who cannot migrate to Amazon Connect due to existing custom solutions or being locked into a specific contact center vendor, AWS CC I Solutions provides APIs and code examples that allow integration with various contact center platforms. These solutions make use of industry-standard file formats and protocols such as Wave files, MP3s, and SIPRO, ensuring compatibility and ease of integration.

Both Amazon Connect and AWS CC I Solutions leverage the power of AWS language AI services such as Amazon Transcribe for speech-to-text conversion and Amazon Comprehend for call analytics and conversational insights. These solutions also make use of Amazon Bedrock, a generative AI platform that allows for call summarization and other advanced tasks.

Overview of Post Call Analytics

Post call analytics is a crucial component of generative AI in contact centers. It involves analyzing customer interactions post-call to derive valuable insights and improve business performance. Using AWS services such as Amazon Transcribe for transcription and Amazon Comprehend for call analytics, organizations can extract important information from customer conversations. Post call analytics provides insights into overall call sentiment, agent performance, emerging business trends, customer complaints, and areas of improvement.

To implement post call analytics, organizations follow a workflow that involves uploading audio files to Amazon S3, triggering a Lambda function, and initiating a step function workflow. This workflow combines various AWS language AI services, including Amazon Transcribe and Amazon Comprehend, to generate call summaries and uncover insights. The data, including transcriptions, insights, and call summaries, is then stored in Amazon DynamoDB and Amazon S3, forming a data lake for future analysis.

Organizations can access the generated insights using the Post Call Analytics (PC A) console, which provides key details such as call metadata, transcribed details, sentiment trends, and more. Additionally, organizations can use AWS services like Amazon Athena to write SQL queries and build aggregated insights, and Amazon QuickSight to Create dashboards for visualization.

New Features of Amazon Transcribe

Amazon Transcribe, an essential component of post call analytics, has recently introduced new features to enhance accuracy and language support. The new multibillion-parameter speech foundation model supports over 100 languages and provides a 30% relative accuracy improvement across all locales. This model is trained using the best supervision approaches and learns from millions of hours of unlabeled audio data, enabling accurate transcription and enhanced readability with improved punctuation and capitalization.

Another exciting feature introduced in Amazon Transcribe is call summarization as part of the transcribed call analytics API. This feature allows users to generate call summaries, identify issues, action items, outcomes, and sentiment with a single API call. It facilitates efficient analysis of customer conversations, saving time and effort. Users can even opt for redaction of sensitive information in the transcript and summary.

These new features empower organizations to make the most out of Amazon Transcribe, providing accurate transcriptions and comprehensive insights from customer interactions.

Principal Financial Group's Journey with PC A

Principal Financial Group, a global investment management leader, embarked on a journey to transform their contact center operations using post call analytics and generative AI. By leveraging PC A, they were able to process over 1 million calls, gaining valuable insights into customer interactions and enhancing their customer experience initiatives.

During their journey, Principal Financial Group partnered with AWS through programs such as the Architect Resident Program and A Data Lab, allowing them to refine and personalize the PC A framework. They were able to extract actionable insights from customer conversations, improve agent productivity, and enhance overall contact center performance.

Principal Financial Group also focused on integrating additional channels like customer email into PC A, enabling a comprehensive view of customer interactions and improving their OmniChannel customer engagement strategy. They also worked on refining their topic hierarchy definition to better capture Relevant business domains and topics.

Looking ahead, Principal Financial Group plans to deploy intelligent agents supported by AWS Q&A Bot and Bedrock. This will enable them to provide personalized and anticipatory customer experiences through chatbot interfaces and advanced knowledge bases. They are also exploring the capabilities of AWS Kendra for efficient keyword search within customer interactions, enhancing their ability to proactively address customer needs.

The journey of Principal Financial Group showcases the power of generative AI and its potential to revolutionize contact center operations, improve customer experience, and drive business growth.

Approach and Deployment Phases

Implementing post call analytics and generative AI in contact centers requires a strategic approach and careful planning. Principal Financial Group adopted a phased deployment strategy to ensure seamless integration and maximum impact. Let's explore the deployment phases they followed.

Technical Deployment

The initial phase focused on technical deployment, including minimum viable products (MVPs) and activities related to transcription. Principal Financial Group ingested data from their engagement center platform (Genesis Cloud) and used AWS Transcribe to generate high-quality transcripts. They then utilized Amazon Comprehend to perform sentiment analysis, topic and intent identification, and personal identifiable information (PII) reduction. This phase also involved reporting enhancements and additions using Amazon QuickSight and AWS Glue.

Topic Hierarchy Definition

The Second phase involved creating a topic hierarchy definition tailored to Principal Financial Group's specific business domains. This definition was crafted in collaboration with their business stakeholders and served as the foundation for their analysis. The topic hierarchy definition was used to cluster and analyze call outcomes, providing deep insights into customer interactions.

Reporting Enhancements and Additions

To augment their reporting capabilities, Principal Financial Group explored additional channels and integrated them into the PC A framework. They began processing email interactions and started incorporating Google Analytics for digital interactions. This allowed them to enhance their topic hierarchy definition and provide a holistic view of customer interactions across multiple channels.

Virtual Assistants and Emerging Theme Detection

Principal Financial Group also embarked on the journey to deploy AWS Lex, an intelligent virtual assistant, based on PC A data. This virtual assistant, infused with AWS Bedrock, used the topic hierarchy definition to address customer queries and provide proactive assistance. They also introduced emerging theme detection to identify important topics and address potential friction points in the customer experience.

The phased approach and deployment strategy allowed Principal Financial Group to gradually harness the power of PC A and generative AI in their contact center operations. With each phase, they gained deeper insights and achieved a comprehensive understanding of their customer interactions.

Improving Customer Experience with PC A

Principal Financial Group's journey exemplifies how PC A and generative AI can be leveraged to enhance customer experience. By adopting a holistic view of customer interactions and understanding their needs, organizations can provide Simplified, personalized, and anticipatory customer experiences. Let's Delve further into the strategies employed by Principal Financial Group to enhance customer experience.

Holistic View of Customer Interactions

To provide a comprehensive perspective on multichannel customer engagement, Principal Financial Group's PC A framework integrated customer email interactions with voice interactions. By analyzing the data from various channels, including voice and email, organizations can gain a deeper understanding of customer behavior and deliver personalized experiences.

The Power of Topics and Intents

Principal Financial Group recognized the significance of topics and intents in connecting different channels and improving customer experience. By analyzing topics and intents across multiple channels, organizations can identify Hidden relationships and gain insights into customer preferences and pain points. This, in turn, enables organizations to provide proactive solutions and build stronger customer relationships.

Omni-Channel Experience

For a seamless customer experience, Principal Financial Group emphasized the importance of an omnichannel approach. By leveraging AWS Neptune, a graph database, they connected voice and email interactions, laying the foundation for further integration of customer surveys, social media interactions, and digital interactions. This omnichannel experience enables organizations to deliver consistent and personalized experiences across various touchpoints.

Roadmap for Future Enhancements

Principal Financial Group, in collaboration with AWS, has laid out an exciting roadmap for future enhancements. They plan to deploy intelligent agents supported by AWS Q&A bot and Bedrock. These intelligent agents, informed by PC A data, will provide personalized and anticipatory customer experiences through chatbot interfaces. Additionally, Principal Financial Group aims to leverage AWS Kendra for advanced keyword search capabilities, further enhancing their ability to address customer needs and improve overall satisfaction.

The strategies employed by Principal Financial Group highlight the potential of PC A and generative AI to revolutionize customer experience and drive business growth.

Conclusion

In this article, we explored the transformative power of generative AI in contact centers. We discussed the key challenges faced by contact centers, the roles of different personas in the contact center ecosystem, and the solutions available to address these challenges. We examined the benefits of implementing generative AI through real-life case studies, showcasing the positive impact on customer experience and business performance. We also provided an overview of Amazon Connect and AWS CC I Solutions as flexible options for organizations looking to leverage generative AI in their contact centers.

Furthermore, we delved into the concept of post-call analytics, highlighting its significance in deriving valuable insights from customer conversations. We explored the new features of Amazon Transcribe that enhance accuracy and language support. To illustrate the practical application of generative AI in contact centers, we followed the journey of Principal Financial Group and how they have leveraged PC A for improved customer experience. Lastly, we discussed the roadmap for future enhancements and the potential of intelligent agents and Q&A bots in contact centers.

By embracing generative AI in contact centers, organizations can enhance customer experience, optimize agent efficiency, and drive business growth in the rapidly evolving digital landscape.