No plan B: Rebuilding CX for the hybrid workforce

Fourteen years ago, when I started Talkdesk, the legacy contact center world was built on complex on-premises servers, multimillion-dollar implementation contracts, and deployment timelines that spanned years. We disrupted the market with a simple promise: Create a call center in 5 minutes. We used the cloud from day one to make the contact center simpler to deploy, operate, and scale.

Now, we’re driving an even bigger transformation.

When generative AI emerged, most of the industry treated it as an add-on, a copilot or drafting assistant bolted onto legacy software. But it became clear to me that helping human agents with copilots was only step one. The technology was moving so fast that AI wasn’t just going to assist work, it was going to execute work. It should be able to reason, act, and resolve complex enterprise work end-to-end.

Eighteen months ago, we made the strategic decision to pivot Talkdesk to Customer Experience Automation (CXA). We saw early on that competitive advantage wouldn’t come from adding more standalone bots or scripts, but from creating an orchestration operating model where AI agents, human employees, data, and workflows operate as one unified workforce.

The market has since validated our strategic conviction. AI can do real work, and CX leaders everywhere are trying to understand what that means for their organizations. Where is AI actually delivering value? What does it take to move beyond isolated use cases? And what are the organizations furthest along doing differently?

That’s why we commissioned this research—The State of Agentic Automation in CX. We want to give the CX community a clear benchmark for where the industry stands today and a practical view of the path forward. We heard from more than 250 CX, IT, operations, and AI leaders across industries and around the world about how they’re putting AI to work in customer experience.

What we found is that AI adoption is moving fast, but the ability to turn it into real outcomes isn’t keeping pace.

The AI gap: Activity is not maturity.

CEOs and boards today aren’t asking for AI activity anymore; they want to see the business impact. Yet, there’s a clear gap across operations. Companies are deploying plenty of AI, but struggling to see real transformation.

The benchmark data clearly reveals the gap. It proves exactly what I hear in my conversations with fellow CEOs, which is a dramatic shift in market urgency. Twelve months ago, there was curiosity surrounding AI. Today, it is no longer a question of if, but how fast.

Our research shows that although 98% of organizations have deployed AI in their customer journey, nearly 80% remain stuck with 10 or fewer isolated automations, struggling to scale value. Here’s a stark reality: only 5% of companies can clearly quantify and report AI’s impact on business outcomes. In the rush to move fast, many companies fell into the trap of buying 10, 20, or 30 standalone point solutions for chatbots, routing, or voice scripts. Point solutions feel quick at first, but create a hidden double cost: you pay once for the software, and a second time in human rework when the bot fails and sends the customer back to an agent who has to start over.

Routing work is not resolving work.

No business wants to manage 30 fragmented tools. It wants one unified platform that handles customer experience needs across the entire business.

The survey data highlights why this platform approach is necessary. While 64% of organizations run specialized AI agents for isolated tasks like billing or identity verification, only 35% have AI capable of maintaining customer context and acting across backend systems to drive true resolution. When bots lack persistent memory and shared context across channels, every handoff forces customers to repeat themselves, turning what should be a fast resolution into a frustrating, disjointed experience.

Routing work between tools is not the same as resolving work. When systems don’t talk to each other, human agents lose up to a third of their workday simply switching between screens and hunting for data. If the operating environment is too fragmented for human employees to work efficiently, AI agents will stall at the same boundaries.

True resolution requires connecting AI agents, human employees, knowledge, customer data, and workflows into one governed operation so customer issues can move from answer to action seamlessly.

The hybrid workforce: Where CXA delivers results.

Just a year ago, there was the belief that AI would completely replace humans in customer service. My view has always been different: the future is a hybrid workforce. Human expertise and AI working together, while elevating human roles to higher-complexity work.

The report data strongly validates this approach. Ninety-nine percent of enterprise leaders agree that a human-AI hybrid workforce is beneficial, and nearly one in five already views AI agents as digital labor rather than software tools. Treating AI agents as digital employees means recognizing that they need the same foundations as human employees: a unified platform, access to clean, integrated data, trusted knowledge at the point of work, and clear governance.

This is already happening in real time across our customer base. For example:

• Humann replaced a labor-intensive, intent-matching chatbot with an agentic AI agent built on Talkdesk CXA, named Hannah. Integrated directly with their order management system and product knowledge, Hannah acts as a member of the team, reasoning, adapting, and knowing precisely when to assist customers independently, providing guidance 24/7 with zero hold times, and when to bring in a human agent.

• Evara Health, a federally qualified health center serving underserved communities, uses Talkdesk CXA to expand access for patients who can least afford to wait. AI agents handle 45% of call volume, slashing wait times from up to an hour down to just 4 to 5 minutes. Because AI handles routine inbound calls, Evara’s staff can reach patients proactively rather than only reacting to incoming requests, a proactive care model the health center could never deliver before.

• InfoPay uses Talkdesk AI to handle repetitive tasks like refunds and account updates—moving away from the old trap of measuring productivity by ticket volume alone. Eliminating that repetitive busywork freed their team to focus on what humans do best: solving complex customer problems, analyzing usage patterns, and delivering real empathy.

The report data mirrors what these customers are experiencing on the ground. CXA leaders, the 15% of organizations that pair agentic AI with cross-system orchestration, are four times more likely to report major CSAT and NPS gains compared to those just one level lower on the maturity curve. Over a third of these leaders autonomously resolve more than 40% of customer issues, while also driving top-line revenue through predictive outreach and personalized retention.

True transformation has no plan B.

Transformation requires conviction and clarity on where the industry is headed. You can’t expect your team or your customers to embrace change if you aren’t completely aligned on the destination. In 2017, long before ChatGPT or generative AI hit the headlines, we started developing an underlying AI orchestration platform. Because we laid that foundation years ago, we were able to embed advanced LLMs directly into our platform at launch, giving our customers immediate, enterprise-grade value.

We didn’t take shortcuts with point solutions, and we didn’t build isolated bots. We built a unified platform that brings together CCaaS for the human workforce and CXA for the AI workforce, all on a shared foundation that connects AI, human employees, data, and workflows across the enterprise.

The competitive moat of the next decade will not be defined by how many AI models or tools a company deploys, but by how seamlessly they’re orchestrated across systems, data, and humans to deliver trusted, measurable business outcomes.

The market is clear, 83% of survey respondents are expecting autonomous AI to only increase over the next two years, with 99% seeing the benefit of a human-AI workforce. The move toward a fully orchestrated hybrid operation, where AI autonomously resolves routine tasks and human employees handle complex issues, is happening now and gaining momentum. Talkdesk provides a platform where people and AI agents collaborate on exceptional experiences.

To give executives a clear roadmap, our report outlines a 5-stage CXA maturity model—providing a clear framework to evaluate where your enterprise sits today, identify the readiness gaps holding you back, and define your next move. The question for leadership is: As you build your human-AI hybrid workforce, which maturity gap will you close first?