The rise of copilot and autopilot in customer support

This article is based on a keynote presentation Neelu gave at the Customer Support & Experience Summit in San Francisco in October 2024.

Customer support is at a crossroads, and AI is the driving force behind this transformation.

I’ve led product growth of complex B2B enterprise SaaS products at some of tech’s biggest companies, which has shown me how businesses grapple with evolving customer expectations. But today, the advent of generative AI has skyrocketed innovation.

Businesses are now rethinking how they engage with customers, streamline processes, and deliver value. AI customer support isn’t just a tool – it’s becoming a fundamental strategy.

Imagine you’re on a flight. The pilot announces two options: “We can use AI to assist me throughout the flight, or I can switch on full autopilot and join you for coffee in the cabin.” Which would you choose?

In this article, I’ll dive into how AI in customer support is transforming every industry, focusing on the balance between “copilot” and “autopilot” functionalities, the rise of AI agents, and the strategies companies that use AI-generated customer support are employing to stay ahead.

The growing demands on customer support

What customers expect

Today’s customers expect more than ever before. A recent Salesforce study revealed some staggering insights:

81% of agents and 73% of mobile workers say customers expect highly personalized interactions.
Yet, over 50% of customers feel that their experience is impersonal.

This isn’t just a set of numbers; it’s a wake-up call for businesses. Customers want meaningful, seamless support – and they want it now. Whether they’re a small B2C consumer or a large B2B client, the expectation for personalized interactions transcends industries. But delivering on this expectation is easier said than done.

The limits of traditional support

For years, customer support has been built on rule-based systems and rigid processes. Companies set up sprawling teams with roles like technical account managers (TAMs), support account managers (SAMs), and customer success managers (CSMs).

These structures aim to cover every touchpoint in the customer journey, but they often create confusion for customers.

When I joined ZoomInfo, I saw this issue firsthand. The first onboarding email we sent to customers included five different contacts – each responsible for a specific issue. Imagine being a customer trying to figure out who to contact for help with a single email, much less remembering those details a month later.

This highlights a critical flaw: traditional customer support puts the burden on the customer to navigate systems. Instead of focusing on their needs, customers are forced to figure out how to work within the company’s framework.

The role of AI in customer support

What is copilot AI?

Copilot AI is designed to work alongside human agents, enhancing their ability to solve problems efficiently and effectively. Think of it as a productivity booster that streamlines processes and supports decision-making. Here are some ways copilot AI works:

It provides agents with real-time cheat sheets, summaries, and relevant insights.
Tools like Salesforce Einstein or Zendesk Copilot use AI to recommend the next steps or suggest pre-written responses.
By automating repetitive tasks, copilot AI allows human agents to focus on complex, high-value interactions.

Despite its benefits, copilot AI is primarily inward-facing. It’s focused on improving internal workflows and doesn’t fundamentally transform the customer experience. The human agent is still at the center of every interaction.

What is autopilot AI?

Autopilot AI represents the next step in the evolution of customer support. Unlike copilot AI, autopilot AI can operate independently, managing entire workflows and resolving issues without human intervention. Here’s what makes autopilot AI revolutionary:

It acts as a unified AI agent, interacting with customers across multiple channels.
It delivers consistent and context-aware responses, no matter the complexity of the inquiry.
It integrates seamlessly with specialized agents to handle niche or technical issues.

Think of autopilot AI as an all-in-one solution that can handle a customer’s journey from start to finish.

Companies like Sierra AI are already paving the way with unified systems that simplify and streamline customer interactions. This doesn’t mean humans are out of the picture – they’re still essential for oversight and escalations – but it does shift a significant portion of the workload to AI.

The challenges of implementation

Case study: Airline chatbots

During a recent flight, I encountered a frustrating issue that highlighted the limitations of current AI systems.

My first flight was delayed, and I wanted to know how canceling one leg of my trip would impact the rest of my itinerary. The airline’s chatbot offered generic, scripted options to reschedule my flight but couldn’t answer my actual question.

After several back-and-forth messages, I typed, “Can I talk to a human?” The bot replied, “I am human.” Frustrated, I ended the chat and spent two hours on the phone trying to resolve the issue.

The outcome? My flight was canceled, and I spent hours fixing a problem that could have been addressed in minutes. This experience shows how rule-based AI systems fall short. They can’t adapt to complex, real-world scenarios.

Case study: Complex SaaS billing

In another instance, a financial services customer faced billing discrepancies across multiple SaaS products.

Resolving the issue required the involvement of support teams, TAMs, and CSMs – and it took three days to escalate the problem to the right product team. The result was a frustrated customer and an inefficient use of internal resources.

Now imagine an AI agent trained to analyze billing data and consumption patterns.

It could have identified the root cause in minutes, providing the customer with a solution quickly and sparing the team a lengthy back-and-forth. This is the promise of autopilot AI: transforming complex, multi-touchpoint processes into seamless experiences.

Advancements in AI technology

Small language models (SLMs)

One of the most promising developments of AI in customer support is the rise of small language models (SLMs). Unlike large, resource-intensive AI models, SLMs are:

Lightweight and capable of operating locally, addressing privacy and security concerns.
Able to deliver context-aware, personalized responses without relying on vast amounts of centralized data.

SLMs make it easier for businesses to implement AI solutions without the logistical challenges of larger models. They’re particularly well-suited for companies looking to scale their AI capabilities incrementally.

Unified AI agents

The idea of a unified AI agent is gaining momentum, and it’s not hard to see why. A single AI agent can:

Act as the primary point of contact for all customer interactions.
Use historical data to anticipate and address customer needs.
Seamlessly transition between different communication channels, from email to chat to phone.

By centralizing the customer experience, unified AI agents eliminate the silos that often frustrate users. Companies like Sierra AI are already proving the value of this approach, delivering consistent, high-quality support at scale.

Reimagining customer support

Looking ahead

The ultimate goal of AI in customer support isn’t just automation—it’s anticipation. Imagine a world where AI agents:

Predict potential issues before they arise.
Provide proactive solutions without the customer needing to ask.
Create experiences that feel personalized, seamless, and effortless.

This shift from transactional to anticipatory support is where the industry is heading. And while it’s still early days, the potential is enormous.

How to get started

If you’re considering implementing AI-based customer support, here are some steps to help you get started:

Identify key pain points: Look for areas where AI can make the biggest impact, whether it’s reducing wait times or improving resolution rates.
Start small: Pilot a single AI tool or solution with a specific customer segment.
Focus on augmentation: Use AI to empower your team, not replace them. The goal is collaboration, not competition.
Measure and iterate: Track metrics like first-contact resolution and customer satisfaction to refine your approach over time.

Game plan

Take stock of your current customer support processes: Are they designed to reduce friction for your customers, or are they inadvertently creating more work for them?
Identify one area where AI can have an immediate impact. Maybe it’s automating routine inquiries or improving personalization in your responses.
Set a goal to pilot a copilot or autopilot AI solution within the next three months. Start small, measure your results, and iterate from there.

Remember, the companies that succeed in the age of AI are the ones willing to experiment, learn, and adapt. The tools are here, the strategies are emerging, and the opportunities are endless. What role will your company play in shaping the future of customer support?

The next step is yours to take. Let’s connect, collaborate, and redefine what’s possible in customer support, one interaction at a time.

FAQs

Will AI create new roles in customer success?

Absolutely. As I shared during my keynote presentation at the Customer Support & Experience Summit, many organizations are already embracing roles like digital customer success managers.

These professionals blend marketing expertise with AI fluency to engage customers in scalable, personalized ways. Companies are also consolidating marketing and CS functions under shared strategies, especially for digital engagement.

What happens to customer support jobs as AI scales?

There’s a fear that AI will replace human agents, but the reality is more nuanced. Klarna’s example shows how automation can reduce reliance on routine tasks while opening opportunities for more strategic roles. Leaders must focus on demonstrating CS’s impact on revenue to secure its place as a growth driver.

How do businesses choose between copilot and autopilot AI?

The decision often depends on maturity. Companies just beginning their AI journey may find immediate value in copilot tools that augment existing teams. More advanced organizations can explore autopilot AI for end-to-end automation. Either way, the key is to pilot, measure, and iterate.

The rise of copilot and autopilot AI in customer support