Analyst Perspective on Agentic AI

What to Expect in 2025-2027

Although artificial intelligence (AI) has been around for decades, and the contact center industry has been leveraging it for years, recent advancements in AI have rapidly accelerated. The latter half of 2024 spouted industry excitement for next-level GenAI - agentic AI, also sometimes called autonomous AI. Agentic AI can act as the glue for orchestrating complex processes across data silos, adapting in real-time, in innumerable applications - both inside and outside of an organization, in concert with other digital agents, bots, robots and humans.

The precursors to agentic AI, while capable, were primarily constrained by rules-based workflows and did not act independently, as in proactively making decisions or carrying out tasks and creating workflows on their own. Before the addition of LLMs, we did what we could to make self-service more human-like in quality and capability. The maturation and use of increasingly more capable LLMs lead us towards a digital workforce that can act autonomously, with little human intervention. Designed with input on the user’s mission or vision, along with context on the issue or situation, they can gather information, analyze, plan, make decisions, and act on behalf of the user and act more like the user.

So, what is agentic AI, and how does it differ from GenAI, ChatGPT, and the rest? This insight is an analyst’s perspective on agentic AI and what businesses and solution providers should expect throughout the 2025-2027 timeframe.