It barely needs stating these days, but there is no shortage of AI agents in the customer experience space.
Indeed, if it were the fashion space, the AI agent would be akin to the Labubu: the must-have, ubiquitous trend that people either have or wish they had.
I bet you didn't expect to read that comparison today.
Unfortunately, that's where the similarities end. Far be it for me to question the power of the Labubu, but I imagine AI agents are slightly more complex.
They certainly cover more ground. There are agents to answer calls, qualify leads, chase incomplete forms, manage claims, recommend products, and pull customer context into an agent’s workspace.
And while all you need to do with a Labubu is clip it onto your chosen accessory, successfully implementing a new AI agent comes with a plethora of obstacles to overcome.
CX leaders must navigate legacy platforms, fragmented data, lengthy CRM migrations, and the uncomfortable question of how much an agent will cost once it starts doing real work.
This is not a new issue; it is a persistent and evolving one, with this week's major announcements suggesting that the latest development is a belief that the obstacle is not always the AI itself.
Editor’s Pick: Genesys Opens an AI Side Door
Genesys has launched Standalone AI Agents, giving enterprises a way to deploy its Agentic Virtual Agent without first moving their wider customer-service environment to Genesys Cloud.
In discussing the news, Tony Bates, Chairman and CEO of Genesys, said:
“Enterprises want to move faster with AI, but they need the flexibility to do it on their own terms without compromising trust and governance.”
For years, the CCaaS market has often treated platform migration as the price of admission for the newest capabilities. Genesys is now offering something of a side door.
Organizations can introduce an AI agent into their existing service environment, keep their established workflows and technology investments, and decide later whether a broader Genesys Cloud move makes sense.
The company says customers can put an agent into production within days, using defined rules and guardrails. When an issue needs a person, the agent can pass the interaction and its context to an employee.
Genesys has also launched Genesys Arena, a builder with more than 50 pre-built Agent Kits for industry use cases. The vendor says teams can create a functional agent in as little as 10 minutes, then tailor its knowledge, integrations, guardrails, and brand persona.
Youssef Nokta, Senior Vice President of AI Forward Deployed Engineering at Genesys, said:
“Genesys Arena gives teams a practical head start, so they can spend less time on the foundational work and more time shaping AI around their customers, business and the results they want to achieve.”
The company is also replacing AI Tokens with Experience Units, its new effort-based pricing model. Interactions are classified as simple, moderate, or complex, with customers paying according to the tools needed to complete the task.
On paper, it sounds cleaner than paying for the underlying AI machinery. Yet customers will still want to see whether “effort” is genuinely easier to forecast once agents start using multiple systems, escalating work, and handling longer conversations.
Talkdesk Wants AI to Remember the Journey
Talkdesk is tackling a related problem with Voyager, a preview product designed to orchestrate customer journeys across days, months, or years.
Voyager coordinates AI agents, employees, and enterprise systems around long-term goals, while maintaining customer context across channels and departments. Its pre-built blueprints span healthcare, retail, and financial services.

