Discover the Potential of Generative AI in Service

Learning Objectives

After completing this unit, you’ll be able to:

• Identify ways generative AI helps diagnose and resolve issues.
  
• Describe how generative AI helps capture and share knowledge.
  
• Explain how sentiment analysis can improve customer service experiences.
  

The Promise of Generative AI

Before starting, if you would like a refresher on the basics of generative AI, including a review of terms like large language models, check out Generative AI Basics.

The generative AI (gen AI) ecosystem is growing at an incredible pace with the release of multiple, highly capable large language models (LLMs). At the same time, many businesses have sprung up to optimize those models and support LLM integration into new and existing applications. This means organizations of all types and sizes have unprecedented access to gen AI. Although there’s a huge interest in putting this technology to use, many organizations want to know how gen AI can improve productivity, inspire creativity, and reduce monotony.

When any new technology emerges, it’s hard to predict its future uses and whether or not it’s easy to use. For example, when researchers for the US government created ARPANET to share information with each other, they had no idea it would eventually give us the internet.

Similarly, how gen AI is used in the next 6 months will look very different from how it’s used 6 years from now. This is a factor of both how quickly the technology is improving and our own creativity in using the tool. It doesn’t help that we often take our ability to perform language-based tasks for granted, so it’s easy to overlook ways to leverage gen AI.

While we can’t predict the future, we can change the way we think about gen AI by looking at examples of how it can be used today. This module explores gen AI from multiple angles. You learn how it can be applied in various parts of an organization, including service, sales, commerce, marketing, and IT. This isn’t a definitive encyclopedia of use cases, but inspiration. As you learn about the promise of gen AI, themes and patterns surface. And before long, you can start to imagine your own variations on those themes.

Supercharge Service with Generative AI

Service organizations handle various language-based tasks. Each day, service reps communicate with customers over the phone, chat, and email. In the process of solving issues, they research and document information. It’s no wonder that good communication skills are sought after for anyone in service organizations. It’s also why gen AI is so effective in this setting.

It doesn’t take long after initial contact with a customer to find a good use for gen AI. It might seem obvious, but you can’t solve a problem if you don’t first understand what it is. Customers might share information to identify the issue, but what they offer is often incomplete or includes irrelevant details. So asking good followup questions is critical.

This is where gen AI comes in. It can use details from prior cases, recent purchases, changelogs of products, and even notes from contemporaneous cases to create clarifying questions. “Is this related to your recent purchase of Widget Pro?” “Have you recently updated your device?”

These questions might be presented to the customer by a chatbot or offered to a service rep by a virtual assistant. Answers lead to additional questions generated in real time. Gen AI can suggest questions highlighting an avenue to explore that otherwise would have been overlooked. With the help of gen AI, you can spend less time identifying an issue, and get right to resolving it.

A gen AI that’s fine-tuned on your knowledge base makes the resolution step faster too. The context that helped identify the issue can also be used to offer solutions, with references to supporting resources. You can even direct the gen AI to explain the reasoning behind a solution. That way, service reps can do a level of quality control, dismissing suggestions they know aren’t helpful. If a solution looks promising, reps can investigate and verify that it’s appropriate.

Gen AI continues to assist by making it easier to share solutions. With the click of a button, service reps can use gen AI to draft a clear, courteous, and relevant response to the customer. This is where gen AI really shines.

You might help a customer who would appreciate a complete explanation of why a problem occurred. Or, you might be helping someone who’s over their head, frustrated, and just wants a quick end to the problem. For either scenario, gen AI can craft a response that best fits the expectations of the audience. This kind of personalization leads to stronger, more trusted relationships. And it’s easy to accomplish when you have gen AI on your team.

But helping the customer find a solution isn’t the end of the story. Service reps need to record details of the case, including a summary of the issue and its resolution. And you guessed it, gen AI simplifies this step too. Generated summaries are concise, scannable, annotated, and tagged with keywords; perfect for other service reps and their managers.

Throughout this example, the service rep chose the best path forward. Gen AI presents options, then the rep uses their best judgment to accept the suggestion or investigate on their own. The hope is for fewer dead-ends, a faster journey, and a happier customer.

But are they happy? It can be hard to know. Surveys are often ignored, and a customer’s attitude isn’t always captured in the case notes. Gen AI offers some solutions to this problem, too. For one, gen AI is great at analyzing sentiment, and can summarize the overall interaction between the customer and service rep. Did the customer show frustration or anger at any point? Did that change over time, for better or for worse? This can help identify high-performing service reps and those who need improvement.

Sentiment analysis can also help route cases to the best agent for the job. For example, a difficult escalated case can be automatically routed to a more senior service rep who excels at defusing situations, while shielding the newer employees from getting out of their depths.

[AI-generated image using DreamStudio at stability.ai with the prompt, “A robot holds a shield in one hand, drawn in the style of a comic book.”]

Finally, gen AI can draft knowledge articles, converting details from one or more resolved cases into data-cleansed content appropriate for a general audience. It could even offer suggestions for pertinent topics based on recent case activity. In minutes you can have an article ready to share with your community on your help portal.

Service organizations are a great place to get a glimpse of the gen AI future. In the next unit you learn what it can do for sales and commerce operations.

Resources

• Help: Einstein Generative AI
• Help: Einstein Generative AI Glossary of Terms
• Trailhead: Generative AI Basics
• Trailhead: Natural Language Processing Basics
• Trailhead: Artificial Intelligence for Customer Service
• Salesforce: Top Generative AI Statistics for 2023