Welcome to the second piece in TheLoops AI in Action recap series. You can read part one of our AI in Action recap here.
AI is reinventing Support and CX operations as we know it. 2025 will be a marquis year for many leaders and teams exploring the role of Agentic AI, digital coworkers and furthering cost containment.
In our continued exploration of AI’s transformative role in customer support and CX, we dive deeper into the strategies, insights, and best practices shared during the “AI in Action” live event on December 4th.
Below, we’ve highlighted key soundbites from our expert panelists. For a comprehensive view, be sure to watch the full discussion.
Our AI in Action leaders covered some critical topics:
- With today’s AI, how should you rebuild your tech stack and processes?
- How can you determine who on your team should be the one to test and refine the first few uses cases?
- In a sea of AI everywhere, what framework can you use to pick the right use cases and providers?
The Promise of AI Platforms: Customizable, Configurable, Impactful
As we pick up from where we left off in our last ‘AI in Action’ recap, we start with what Greg Giletto of Bloomreach and Genady Rashkovan of Tricentis discussed during their session, “Breaking Out Of A Single Solution Mindset.”
Traditionally, Support and CX teams relied an array of point solutions based on the organization’s size, complexity, and industry. On average, small to medium sized CX teams might have had 5 to 10 tools while larger, enterprise teams could have anywhere from 15 to 30 as they often require more specialized and integrated tools to manage complex workflows and high ticket volumes.
AI platforms, on the other hand, replace the need for multiple tech stacks. They are a much more efficient, beneficial and versatile way to consolidate tools, minimize data silos and ensure an overall increase in CX efficiency and workflows.
“When we think about AI and support data and providing access to multiple teams, we need an AI platform vs multiple point solutions to create those feedback loops. Finding a solution for VOC and a separate solution for QA then one for Copilot and maintaining those systems, thats a full time job in and of itself. Leveraging a platform is much more agile, strategic and seamless,” Greg shared.
Watch this quick clip here where Greg further expands upon the benefits of AI platforms that keep multiple teams in post-sales ‘within the loop’—Support, Product, Success.
When it comes to AI and CX impact overall, Genady also shared that Support needs to think bigger than chatbots, decision trees and single solutions. “The approach to AI now is thinking about an AI engine with an orchestration layer that is working and feeding everything back to the bigger engine with hierarchy. AI should be helping you understand and improve the customer journey, and find ‘quiet customers’ who may never escalate a case because they aren’t even using your solution.”
He specifically cites Copilot, Product Feedback Analysis and AutoQA as great examples of this.
- Copilot addresses the biggest pain of helping your agents immediately know how to unblock customers by way of suggested next steps, historical summaries and even recommended knowledge
- Product feedback analysis addresses retention issues and helps your product team have measurable, real-time customer data based on revenue at risk and recurring trends
- AutoQA helps CX leaders quickly put a stop to errors and missteps causing low CSAT and instead, pinpoint areas where agents can receive immediate coaching then inform new and improved training
“You need AI to help you monitor the value you’ve sold, is the customer getting that? And then you want AI that helps you monitor the delivery,” Genady shared.
When it comes to breaking through the noise, and determining the right AI to start with, Genady also suggested this simple but clarifying exercise: “Think about the problems you are trying to solve for the business then go from there.” He expanded upon this further below.
Lastly, Greg also shared that it’s key to work with a partner who doesn’t have a “one-size-fits-all” approach to AI.
“The biggest benefit I’ve seen with working with TheLoops is the team’s willingness to work with us to really ensure the use cases we’ve selected for our unique business needs and data are providing maximum value, there’s a lot of tweaking, configuration and customization that they help us with.”
Our next session was the perfect segue into a topic many leaders are examining: The AI Adoption Trifecta featuring valuable input from Mariena Quintanilla and Daniel Rose.
As businesses navigate the complexities of modern CX, aligning your people, process, and technology is the cornerstone of success. With that said, having the proper expectations around how fast AI will work, knowing how to improve its accuracy with reinforcement learning and considering the best ways to drive adoption should not be an afterthought.
“The first perception of AI is everything,” Daniel shared. “If your team doesn’t understand which use cases you’re deploying, that it (AI) won’t work perfectly on day one or, how you will measure value, that will create a slew of challenges from limiting agent adoption to awkward conversations with your finance team when it comes time for renewal.”
Mariena also emphasizes that Support leaders need to know two key things to manage initial expectations of AI working like instantly like an “on and off” switch:
- AI accuracy and implementation does require experimentation. “I come from an engineering background and for developers, it’s very common, especially if you’re working on a machine learning or an AI project to know that you’re going to be iterating and you’re going to build a first round of a model, you’re going to review the results then, you’re going to adjust things, and you’re going try again. Very similarly, when you’re deploying AI products in your CX organization, it behooves us as leaders to manage expectations with our team and manage up within our organization that the first thing we’re going to do is get some baseline information and that will help us understand where we need to improve.”
- Select the people on your team to deploy AI who embrace being iterative and who thrive on experimentation. “Identify the people within your team that you think can handle seeing things a little bit messy. When you’re rolling these tools out, you have to identify those people on your team that can help build this, right? And build by experimenting. AI should be lead by the ones that are comfortable seeing things messy, and they can see your vision and what you’re trying to accomplish. They can see where you’re trying to go and help you get there by providing constructive feedback. If the person you choose to spearhead this is going to come in and just talk about how crappy a tool is and spread that throughout the organization, that’s going to tank your progress.”
Watch this quick clip here where Mariena and Daniel elaborate.
As we wind this recap down, Daniel does suggest leaders also remember: “AI tends to as good as the quality of the data that’s behind it. Understanding the data quality that you are working with to make these assumptions and drive results is critical.” Last but not least, think of training AI as you would a new hire. It will take a few weeks to get up to speed, but once it does, all bets are off.
We’ll be back next week with recap 3 rounding out our final talk with Kartik Yegneshwar, VP of Global Support and Technical Success at Gainsight. During his session, we covered AI use case selection from the lens of the lifecycle of a ticket. Kartik also expands upon driving AI adoption across your team with some gamification ideas and more.
Curious to see how TheLoops platform can help you and your team in 2025 and beyond? Set up a call to discuss your needs here.
