Agentic AI journey: your guide to a customer-centric experience

The pressure is on. Customers expect ever richer, more personalized and responsive engagement across all channels, from contact centers to social media. 

Here’s where AI is game-changing. Busy companies are using it to reduce time-to-serve, adapt quickly to shifting demands, and lighten the load on customer service teams – all while increasing consistency and quality across every customer touchpoint.

But AI is moving fast. You may be feeling a bit overwhelmed by the dizzying speed of its advance, the variety of GenAI tools, and how to integrate them into an existing environment. Or you may be wondering why the benefits of AI are not yet materializing for your business.

Firstly, let’s break down the recent evolution of GenAI into four key stages (each of which integrates the capabilities of the previous stage).

Stage 1: Text generation. This is the ability to autonomously produce human-like written content. At Konecta, we’ve designed and implemented lots of digital assistants that help human agents by autonomously drafting, summarizing and translating text. While simple, this is proven to help agents work faster, while also enhancing fluency in multiple languages.

Stage 2: Retrieval. These tools can consult external sources and tap into diverse knowledge bases to pinpoint and integrate relevant information. We’ve developed examples in multiple sectors, saving agents significant time and producing highly accurate responses to specialized customer queries.

Stage 3: Short- and long-term memory. AI agents can remember the context of previous conversations, hence building continuity and coherence of responses over time. We’ve implemented real-time AI copilots that equip human agents with personalized scripts, best-next-step suggestions and alerts for resolving customer queries comprehensively and fast.

Stage 4: Integrated agent architecture. These are, in essence, intelligent AI agents that autonomously make decisions and execute processes, using other existing tools and integrating structured and unstructured data. Recent examples we’ve worked on include an intelligent voice AI agent in retail that can process half of all orders, escalating only the more complex to human beings. 

 So, if you’re a customer-centric business, what does this evolution mean in practical terms? You might leverage any or all these capability types – depending on your customers, your environment – and what you want to achieve.

This aim is to build what’s called ‘composable architecture’ – a future-proof way to devise and implement a modern digital ecosystem. It means designing and integrating GenAI capabilities as modular components, connected by application programming interfaces (APIs). Changes can be made to individual modules without impacting the whole ecosystem, so you can improve and adapt your customer operations with greater speed, flexibility and scalability.

Wherever you are on your AI journey, a great place to start, restart, or take stock of your progress is to carry out a holistic AI maturity assessment. This should encompass all relevant business functions, as well as IT, with input from leaders together with customer service agents (who tend to have great insights when it comes to their customers’ experience). 
Remember: designing for AI means designing for dialogue – so co-creating solutions with the people who live these experiences day to day is what drives speed and maximizes value.

An AI maturity assessment should include:

• Pain points and key metrics. Where are the biggest barriers, problems and sticking points in your current customer journey? What would success look like? Are there KPIs that define your goals?
• Business processes. Which processes and touchpoints are key? What can be automated or redesigned? Where could knowledge be better leveraged for customer experience and operational efficiency?
• Data. What data is required and where is it? How is it classified? Is data complete, secured and AI-ready? Are there data siloes hindering progress?
• Technology. Is the infrastructure on-premise, hybrid or cloud-based? Are there API limitations that could impact AI deployment? 
• Risk and compliance. Which regulations and industry standards must be met? What are the biases and ethical concerns related to AI-driven decision-making?
• Business readiness. Who and where are the key users? How will their role change? How can they be empowered and upskilled?

The answers to these questions will, of course, depend on the specifics of your business. They will inform and shape your transformation roadmap as your technology and business landscapes evolve.

In future blogs, we will unpack the critical success factors for AI-powered transformation. In the meantime, here are three important lessons from our recent experience.

1. Establish a robust governance framework so your AI initiatives remain compliant, ethical and transparent (the world of ‘responsible AI’). Addressing and mitigating bias – together with ensuring tight security and privacy – is vital to safeguard reputations and ensure sustainable AI-driven growth.
2. Identify the right use cases (evaluated on readiness and returns), together with clear metrics to assess the impact of AI, such as response/resolution times, customer satisfaction scores, and operating cost reductions.
3. Don’t be afraid to experiment. Experience in leading companies underlines the importance of a ‘fail fast, learn fast’ mindset. Launching pilots enables you to test capabilities in controlled environments, while developing best practice and refining integration of AI with your legacy systems. 

The Fifth Industrial Revolution will be defined by ever closer and interdependent human-machine collaboration. Paradoxically, as AI becomes more sophisticated, the customer experience becomes more human. That’s because AI empowers human agents to personalize responses and interact more empathetically and personally, thereby enhancing the perception of care and attention towards individual customers.

While the future can never be certain, your next step can be if it’s based on a clear-eyed view of your business today. People remain at the heart of AI innovation: involving yours in your AI assessment could unlock potential that will power your ambitions in the years ahead. 

-
With the contribution of Oscar Verge, Chief AI Deployment Officer, and Diana Catalina Velasquez, Head of AI Deployment for LATAM.