At the Genesys CX Tour India 2026, Aditya Garg, Head of Solution Consulting, India, discussed what it means for a customer experience (CX) system to act autonomously. The audience had several interesting questions around the role of artificial intelligence (AI) in CX, but four of them stood out.

Read on as we break down those questions and showcase what they mean for organizations that are trying to figure out where agentic AI fits in their CX operations.

Missed the event? Watch the session recordings to hear industry leaders share practical strategies for AI-powered experience orchestration. https://www.genesys.com/en-sg/events/cx-tour-india-2026

The Shift in Customer Experience

Many organizations that say they are using AI in customer experience are just automating a few tasks. These “AI” tools respond to customer queries, route them to the appropriate agent, and escalate them if the resolution is complex.

At the Genesys CX Tour 2026, there were several discussions on customer experience platforms that can adapt to real-world complexity. Here are four agentic AI and CX questions (with answers) from the audience:

Question 1: How is traditional automation different from agentic AI in customer experience?

The real difference between traditional automation and agentic AI is about what the system is designed to do when it encounters a situation that was not anticipated at build time.

  • Rule-based vs. autonomous: Traditional automation systems execute tasks in steps based on a set of pre-defined rules. An agentic system AI receives an objective and decides which steps to take, in what order, and using which tools – based on the context.
  • Fixed workflows vs. adaptive behavior: Traditional systems require developer intervention every time a new scenario emerges. Agentic systems learn from interactions and adjust over time. They respond to new queries and improve outcomes without constant manual updates to the underlying logic.
  • Completing tasks vs. achieving outcomes: A bot that can confirm a flight booking has done its job. An agentic system, on the other hand, can detect if a customer is at risk of abandoning a purchase, proactively surface a better option, and close the interaction without a transfer.

Question 2: How does the shift from automation to autonomy change the way AI systems operate in CX?

Automation assumes the organization has mapped every scenario worth handling. Autonomy assumes the customer’s situation is unpredictable and gives the system the ability to decide what to do without a predetermined path for every case.

  • Objective-driven rather than workflow-driven: An autonomous system gets a goal and determines how to pursue it. It removes the ceiling on what a system can handle without a rebuild every time the business adds a new product line or changes a policy.
  • Initiative rather than reaction: Agentic AI systems monitor customers’ behavior on a website and intervene before they ask for help. They can read disengagement signals and offer assistance at the right time.
  • Cross-agent collaboration: A single agentic interaction can span multiple agents working across identity verification, inventory checks, appointment scheduling, and a CRM update. They complete defined tasks in sequence without a human handoff between each step.

Question 3: What is the difference between Large Language Models (LLMs) and Large Action Models (LAMs)?

Large language models and Large Action Models (LAMs) are both described as AI, both are involved in agentic systems, and both are frequently mentioned in vendor presentations without a clear explanation of what each one actually does. But the difference is simpler than it sounds.

  • Large language models handle understanding: LLMs are trained on articles and conversations; so, they become very capable at interpreting customer intent and generating a relevant response.
  • Large action models handle execution: LAMs are trained on synthetic datasets designed around task performance. Their job is to determine what action the system should take and execute it correctly.
  • Together they enable action and execution: An LLM without an LAM produces good conversation but no outcome. A LAM without an LLM struggles to handle the variation and ambiguity that real customers bring to every interaction. Agentic AI requires both working in sequence, within a system that connects understanding to action reliably and at scale.

Question 4: What are the key components of the Genesys AI framework for agentic AI?

The audience wanted specifics on how the Genesys AI framework for agentic AI is built. Aditya walked through four layers, each addressing a distinct requirement for enterprise-grade agentic deployment.

  • Intelligence layer: This is where organizations design and orchestrate journeys using tools like AI Studio, AI Guides, and AI Skills. It handles goal definition, agent configuration, and the sequencing of steps across channels.
  • Execution layer: At this layer, thinking, planning, and acting come together. The execution layer gets the goal, decides the path forward, and carries it out, using the relevant systems and tools.
  • Trust layer: This layer defines what the system is allowed to do, under what conditions, and what kind of triggers should escalate issues to a human agent.
  • Models layer: Genesys supports proprietary models, open-source options, and bring-your-own model configurations – all under a single framework. If organizations have already invested in a specific model for a specific use case, the framework can still provide orchestration across the full journey.

Narayana Health, a leading healthcare provider in India, chose the Genesys Cloud AI-Powered Experience Orchestration platform to help transform its end-to-end experiences. The result: Faster response times, improved customer service, and a 15% reduction in AHT

Read the full case study → https://www.genesys.com/en-sg/customer-stories/narayana-health

Mastering Orchestrated Autonomy in CX with Genesys

At the Genesys CX Tour 2026, the audience showed real curiosity and energy. People asked lots of questions while talking with the speakers. Many of these questions circled back to the same underlying issue: organizations have already put AI in place, but only in pieces, without a clear picture of how those pieces actually connect across the entire customer journey.

What makes agentic AI genuinely valuable is its ability to orchestrate customer journeys, right from the first contact, all the way through to resolution. That was the shift Genesys laid out at the CX Tour.

If you are looking to build your CX on an agentic AI foundation, here’s how we can help!

Missed the sessions live? Watch the recordings: https://www.genesys.com/en-sg/events/cx-tour-india-2026

FAQs

What makes agentic AI different from a regular chatbot?

Agentic AI focuses on goals and adapts to context, while a chatbot follows fixed scripts and predefined decision trees.

Do organizations need to replace their existing AI tools to adopt agentic AI?

No. Platforms like Genesys support bring-your-own model options and integrate agentic capabilities alongside existing tools.

What role do guardrails play in the deployment of agentic AI?

Guardrails define the boundaries within which an agentic system can act, ensuring it stays compliant, accurate, and within approved parameters.