Conversational AI: Automation with a Human Touch
June 2025
The conversational artificial intelligence (CAI) self-service and agent augmentation sector is experiencing brisk growth due to the convergence of rapidly improving technology and socio/economic events. Agentic AI is currently the CAI segment’s favorite technology as vendors invest to enhance these solutions’ autonomous decision-making, automation capabilities, and effectiveness. The rate at which many of these applications are maturing is both exciting and concerning, as enterprises ask CAI vendors to deliver self-service features that have never before been possible. The timing for these innovations is ideal as businesses are actively looking for ways to reduce operating costs, and replacing live agents with bots is top of mind for many executives.
CAI’s Mission
CAI vendors are on a mission to deliver solutions that help enterprises enhance their customer experience (CX). Generative AI (GenAI), large language models (LLMs), and most recently agentic AI enable the automation of many use cases that organizations were previously unable or unwilling to offer via self-service. Healthcare, travel and hospitality, retail, insurance, financial services, logistics/transportation, and on-demand services are a few verticals quickly adopting CAI solutions to automate a growing number of tasks on both an inbound and outbound basis. While these applications cannot currently handle all inquiries and transactions, they are advancing daily as their technology, intelligence, and decision-making skills improve.
The Transformative Power of CAI
In the past year, contact center communications have changed in a meaningful and lasting manner, which is positive for consumers. The old rules of self-service have been tossed out, replaced by CAI solutions which enable many types of conversations in voice and digital channels including between bots and customers, bots and bots, or simultaneously in multiple channels. The volume of calls coming into contact centers is growing again as CAI solutions deliver more human-like voice interactions, increasing customers’ willingness to use a bot if it is helpful, effective, and will transfer them to a live agent when necessary.
CAI solutions are transforming contact center servicing dynamics by providing real-time guidance and automation that help agents during and after customer conversations. These capabilities enable employees to spend their time engaging with consumers and solving their issues instead of looking for answers, updating customer relationship management (CRM) solutions, summarizing interactions, and initiating follow-up tasks, etc. Intelligent CAI solutions are also proving to be effective tools for proactive customer care and other outreach, delivering significant benefits to outbound activities. These innovative and flexible CAI applications utilize agentic AI together with GenAI to make appropriate offers to customers and decide when (or if) to bring a human into the discussion. When transferring the conversation to their live counterparts, these smart solutions provide context along with the interaction, eliminating the need for customers to start over.
Navigating CAIs Growing Pains
The potential benefits of CAI are great, but this sector is facing challenges commonly experienced in new IT sectors. Due to these solutions’ immaturity and their rapidly evolving underlying technologies, the majority of CAI implementations are taking more time than expected, frequently six to nine months, as compared to the three to four months vendors are indicating. This challenge is exacerbated by the limited availability of AI experts, both in the companies that develop and implement these solutions and in the enterprises where they are implemented. Another challenge is the high price of CAI solutions, which is to be expected with new technology. DMG has started to see prices drop as vendors compete aggressively for business. In the meantime, some enterprises have opted to supplement their existing interactive voice response (IVR) systems instead of replacing the entire solution, to quickly realize CAI benefits. The CAI market will achieve its goals, and due to the inherent capabilities of AI, it will be faster than it would have been as recently as a year ago.
Final Thoughts
The benefits of agentic and GenAI-enabled CAI solutions are already significant, even though these are in their early days. DMG recommends that companies identify and assess the ability of CAI to augment agents and fully automate customer inquiries and transactions to determine the specific ways they can enhance their operating environment. While these solutions are not yet eliminating the need for live agents in most companies, DMG expects them to make major contributions to increasing automation rates by the end of 2027.