Nearshore Americas

The Breakdown: It May Not Seem Obvious, But AI is Already Rewiring Global CX

BPO providers find themselves in a mad scramble to properly understand and implement the latest AI tools, with generative AI being at the forefront of their priorities. 

The state of things: For months, BPO vendors have been experimenting with the latest AI tools, such as ChatGPT. Results are already taking the form of applications that, though still in a preliminary state, provide a glimpse of what business process services could look like in the near future.

  • Teleperformance has been plugging its in-house GPT solution (TP-GPT) to the operations of its CX agents. 
    • The French company claimed in a financial presentation that AI  reduced call length by nearly 40% (2:44 minutes on average, with major reductions in times of the intent discovery and fulfillment phases) and cut-down agent errors by 90%. 
  • Accenture reported to its investors that it has been working with a multinational bank to “transform how it manages high volumes of post-trade processing emails every day […] leveraging a generative AI solution as it is built to understand the context of emails with high accuracy”.
  • Concentrix has been using generative AI to improve its chatbot, advisor assistance and business intelligence solutions. 

Sure, but: Generative AI is not a simple plug-n-play technology. As CX providers experiment with the tools, they’re coming to realize that they have limited applications for business process services, at least in their current state.

  • Teleperformance classified potential generative AI applications for CX into the “easy” ones (tier-1 support, chat, e-mail, rule-based back-office processing, non-critical translation) and the “difficult” ones (complex customer support, sales, collections, trust & safety, critical translation, judgment-based back-office processes). 
  • Monti Becker Kelly, Webhelp’s recently appointed SVP of Global Accounts, told NSAM that one of the most common mistakes clients make when implementing AI solutions is trying to “apply the first level of AI into a complex experience” or trying to use AI solutions that attempt to personalize the experience but fail.

CX vendors –and many other businesses– are coming to grips with what Joe Procopio –one of the minds behind the first natural language generation (NLG) tool to make it to market– realized in his early experiments with AI: “Great tech, lousy application.” 

What about the workforce?: A recent report from Goldman Sachs estimates that “roughly two-thirds of current jobs are exposed to some degree of AI automation, and generative AI could substitute up to one-fourth of current work.”

  • The highest levels of exposure are found among administrative and legal professions, with physically-intensive jobs, like construction and maintenance, being less threatened. 

The paper echoes Teleperformance’s estimations. The company sees (potentially) 20%-30% of its volumes being automated over three years. 

Hold your horses: Goldman’s own report points out that job replacement could be “offset by creation of new jobs and the emergence of new occupations.” 

  • The paper underscores the fact that  60% of today’s workers are employed in occupations that did not exist in 1940, according to the US Census Bureau,
  • When asked about the future of agents in AI-powered CX providers, Webhelp’s Becker Kelly spoke of “advisories [that] should become problem solvers, the white-gloved concierge”.
  • Accenture CEO Julie Sweet described generative AI –”which we’re super excited about”– as a “copilot to human beings.”
    • “The entire process has to be changed in order to make that work. You’ve got to upskill the people and you have to be able to do all of that in a very responsible way,” she said during the latest company conference call.

NSAM’s take: For all the hype around the potential for generative AI in the BPO sector, we’ve only seen what amounts to baby steps.

Applications remain simple and, in some cases, clunky. Consumers have been raging for a while against automated calls and chatbots in customer service.

Sign up for our Nearshore Americas newsletter:

Some might be disappointed with the current results of AI, which remain far from the sci-fi scenarios heralded by many of its strongest supporters. Nevertheless, there’s lots of lab work and experimentation being done.

We expect things to pick up quickly, as scary as that might sound to some. Even then, don’t assume the world will be dragged overnight into an Isaac Asimov (or Philip K. Dick?) novel.

About The Breakdown

“The Breakdown” is the newest feature in NSAM’s portfolio of market intelligence deliverables. It  is designed to give readers a quick, probing look at critical issues in the Nearshore market. Each Breakdown will end with NSAM’s take on the matter, allowing us to leverage our expertise and add our own unique value.

We’re excited about this new product. Share comments and suggestions: Chief Editor, Cesar Cantu (

Cesar Cantu

Cesar is the Managing Editor of Nearshore Americas. He's a journalist based in Mexico City, with experience covering foreign trade policy, agribusiness and the food industry in Mexico and Latin America.

Add comment