How to Use Generative AI in Customer Service

TL;DR

Generative AI can improve customer service through richer self-service, real-time assistance for human agents, and deeper analysis of contact center conversations. Effective deployment starts by defining the intended experience, understanding customers, selecting service channels and supporting tools, then designing an operating model that aligns technology with the broader service strategy.

Transcript

• [Auto-Attendant] All are currently busy. and your call will be answered by the next available representative. Today's customer hates When I call up a brand I expect to get my and to my satisfaction. In the world of customer service, the challenge is to stay a step ahead of customer expectations and ensure that every brand interaction is not just ... Read More

Key Insights

• Generative AI improves customer service in three principal areas: self-service, assistance for human agents, and contact center operations. These applications address both sides of the service equation by creating more seamless customer interactions while increasing the productivity and consistency of the people supporting them.
• Traditional customer service technology is often fragmented across interactive voice response, agent-assist systems, robotic process automation, and chatbots. Generative AI creates an opportunity to connect interactions more smoothly across channels, including allowing a request that begins through one channel to be completed through another.
• Generative AI makes virtual agents more flexible than predetermined dialogue trees. Large language models can recognize, classify, and create sophisticated text and speech, producing richer experiences that are more resilient to conversational variation than journeys based entirely on handcrafted flows.
• Agent assistance works by retrieving and presenting relevant information from knowledge bases during customer interactions. This reduces the time agents spend searching for answers, helps them resolve queries faster, and can also support field service workers who need faster and more accurate troubleshooting guidance.
• AI-assisted writing gives agents contextual drafts that remain subject to human review and editing. The transcript specifically identifies email responses as a practical use case and states that AI-augmented emails have shown higher engagement, while preserving an agent's role in approving the final communication.
• Conversation analysis can reveal patterns across every call in a large contact center. Generative AI can help leaders understand why agents struggle with particular call types, capture granular feedback about products or services, identify problems faster, and alert service, product, or marketing teams when action is needed.
• Automated after-call documentation saves time by transcribing conversations in real time and generating draft summaries of the discussion and actions taken. Agents can edit the drafts before completion, improving consistency while reducing periods when they are unavailable to answer another customer call.
• Customer service transformation follows a five-step framework: define the intended experience, understand customer demographics and needs, decide which channels to support, choose the tools and platform for those channels, and design an operating model that delivers the original service strategy.

Questions & Answers

Q: How can generative AI improve customer service?

Generative AI can improve customer service through three connected applications: richer self-service, assistance for human and field service agents, and analysis of contact center operations. It can produce flexible conversational experiences, retrieve knowledge, draft contextual responses, transcribe calls, create summaries, and examine conversations at scale. Together, these uses can shorten resolution work, increase consistency, reveal service problems, and support seamless experiences across channels.

Q: How does generative AI improve customer self-service?

Generative AI improves self-service by moving virtual agents beyond rigid, predetermined journeys. Traditional chatbot flows require teams to analyze customer intent and handcraft dialogue paths. Large language models can recognize, classify, and generate sophisticated text and speech, making interactions richer and more resilient to conversational variation. New tooling can also turn a process specialist's description of a desired journey into generated dialogue flows.

Q: How can generative AI assist customer service agents?

Generative AI can assist agents by finding relevant information in knowledge bases and presenting it during customer interactions. This reduces search time and helps agents resolve queries more quickly. It can also draft contextual email responses for agents to review and edit, while similar tools can help field service workers troubleshoot problems faster and more accurately. These capabilities let employees focus more attention on customers.

Q: How can AI reduce customer service after-call work?

AI can transcribe a customer conversation in real time and generate a draft summary covering the discussion and actions taken. The agent then reviews, edits, and completes that draft. This approach reduces the time spent manually documenting each call, improves consistency in the details captured, and shortens the period when an agent is unavailable to handle a new customer request.

Q: How does generative AI help contact center operations?

Generative AI helps operations teams examine conversations across contact centers with thousands or even tens of thousands of agents. It can identify how and why agents struggle with particular call types and extract detailed feedback about products or services. These insights allow service leaders to detect and resolve problems faster or alert product and marketing teams when remedial action is required.

Q: What are the five steps for AI customer service transformation?

The five-step process begins by defining the customer experience the organization wants to deliver. The business then studies its customers, including their demographics and needs, selects the channels through which it will serve them, and chooses the supporting tools and platform, such as cloud-based or on-premises options. Finally, it designs an operating model that aligns execution with the original service strategy.

Q: Why should companies use AI across customer service channels?

Customers expect seamless answers across channels and throughout acquisition, service, and retention. AI can support interactions regardless of the channel used, allowing a request that begins in one place to be completed in another. When combined with traditional enterprise capabilities, it can also support proactive outreach intended to prevent problems or address them earlier, creating a more coordinated customer experience.

Q: What customer service results did the US Veterans Affairs example show?

The US Veterans Affairs work described in the transcript began in 2019 and applied analytics and automation to accelerate claim creation and responses. The operation handled 3 million packets spanning 280 document types and 24 distinct mail processes, achieved 100% automation of mail intake, and processed 220,000 items. The effort later added sophisticated AI to help claims adjudicators make decisions faster.

Summary & Key Takeaways

• Generative AI expands customer service automation beyond predetermined chatbot journeys. Large language models can recognize, classify, and generate sophisticated text and speech, enabling more natural self-service interactions. They can also help process specialists create dialogue flows by describing desired journeys, reducing the need to handcraft every conversational path.

• Human agents can use generative AI to retrieve relevant knowledge, troubleshoot issues, and draft contextual email responses for review and editing. Real-time transcription and automatically generated call summaries can also reduce after-call documentation. These capabilities shorten resolution work, improve consistency, and allow agents to serve more customers during their shifts.

• Contact center leaders can analyze conversations at scale to identify why agents struggle with certain calls, capture detailed product or service feedback, and detect emerging problems. Successful transformation follows five steps: define the target experience, understand customers, choose channels, select supporting tools and platforms, and design an operating model aligned with the service strategy.