Contact centers are successfully ramping up their adoption of AI agents to fully automate customer interactions. In fact, 68% of companies using AI had increased their spending on AI triage agents in 2024, according to Metrigy’s AI for Business Success 2024-25 global study of 697 companies.

As generative AI has drastically improved the success of these AI agents, they will ultimately be the front door to any company, driven largely by cost savings and speed of resolution. So rather than have a human or a robotic and slow IVR answer calls coming into any area of a company, a more sophisticated agent can take over that role.

Now, companies of any size can enable a virtual receptionist to their RingCentral phone system with RingCentral’s AI Receptionist (AIR), announced on February 20. Like a virtual AI agent in the contact center, AIR uses generative AI to answer questions fully or route calls appropriately to the fastest resolution. 

Voice remains preferred communications method

A movement to this type of product makes sense, given 77% of all interactions with a company either start in or are escalated to voice, according to Metrigy’s Customer Experience Optimization 2024-25 global study of IT and CS leaders from 544 companies. In fact, 72.1% of consumers cite voice as a channel they use to interact with companies—by far, the highest of any other method. It’s also their preferred method of contacting companies and the fastest method for resolution among the highest percentage of consumers, according to Metrigy’s Customer Experience Optimization – Consumer Views 2024-25 North American study of 502 consumers.

Couple the desire for voice communications with the advanced functionality of generative AI, and companies will start seeing some big improvements to their business success metrics. Cost reduction is an obvious improvement. But a more interesting one is the potential for more sales. For example, when people call a company, an AI receptionist can ask a few lead qualification questions, and based on the answers, transfer the call to a salesperson.

The majority of CX leaders (54.7%) say generative AI has the potential to fully resolve between 26% and 75% of all customer interactions. In fact, the quality of automated AI agents improved drastically during 2024, with 81.6% saying speed of answers and 77.7% saying accuracy of answers improved. What’s more, 74.9% of companies using AI agents reported an improvement in the ability to fully automate interactions. Now, about 22% of all interactions coming into a company are fully automated. Expanding this type of capability to any company phone number, and not just a contact center, will increase the automation.

RingCentral says early customers of AIR have seen solid improvements. For example, a security company saw automated resolution to 50% of its inbound calls, while a healthcare company reduced average answer times from 12 seconds to 0 seconds. Another benefit is capturing leads that would have been lost due to unanswered calls.

RingCentral AIR capabilities

AIR is embedded into RingCentral’s UCaaS platform, which makes it fairly straightforward for any size company using RingCentral’s phone system to deploy this automation. Often, similar AI triage agents implemented in a contact center are more complex point solutions with separate management infrastructure.

RingCentral’s President and Chief Operating Officer Kira Makagon says AIR marks the company’s foray into agentic AI. A rapidly emerging area of AI, agentics combines the words “agent” and “analytics.” The term generally refers to a framework on top of generative AI that uses large language models to make decisions and take actions without human involvement. 

The virtual receptionist does just that. It assesses the caller’s issue or need, and takes action based on data from a company’s website and/or documents in its knowledge base. If AIR can answer the question, such as those about the company’s services, pricing information, or business hours, it will. Otherwise, it will route to the best possible person or department—all based on natural language conversations, vs. a decision tree of “press 1 to be connected to service,” and so on. Once it’s implemented, companies can run tests using voice or text.

A few other key features of RingCentral’s AIR are its ability to deliver information from the conversation via SMS and to deliver transcripts for every call—vital to leveraging conversation analytics that help companies manage call volumes and trends.

Moving forward, RingCentral will add the ability for customers to schedule appointments right with the AI receptionist. The product supports English now, but RingCentral plans to add Spanish at the end of Q1, with other language support coming afterwards. 

AIR is available now to select U.S. businesses, and RingCentral is offering a one-month, 100-minute free trial. Though RingCentral hasn’t announced pricing, John Finch, VP of product marketing, says he is expecting “disruptive pricing models with pay-as-you-go bundled packages.” A broader rollout will commence in second quarter.

Originally published Feb 24, 2025