From Pilot Projects to Majority of Resolutions
AI customer support agents are rapidly shifting from side experiments to frontline operators. HubSpot’s Customer Agent now autonomously resolves 70% of support conversations, up from just 20% a year earlier and five points higher than last quarter. Some customers are already clearing 90% autonomous resolution rates, with Synergent reportedly hitting 85%. That pace of improvement is unusual in contact center automation, where incremental gains have historically been the norm. The product has surpassed 9,000 customers and now consumes over half of all AI credits on the HubSpot platform, well ahead of its prospecting and data-focused counterparts. Executives highlight two core use cases: handling after-hours and weekend support, and taking on tier-one tickets so human agents can focus on complex issues. As underlying models improve, HubSpot expects Customer Agent to move beyond basic deflection into higher-tier support.
HubSpot Turns CRM into an AI Agent Platform
Behind the resolution numbers is a broader architectural shift: turning HubSpot’s CRM into a platform where AI agents both run on and run the system. By expanding its public APIs and MCP server, HubSpot is enabling agents to access and act on unified customer data spanning marketing, sales, and support. This unified data fabric is critical for AI customer support agents that need context to resolve issues rather than simply route or deflect them. The company has also extended Customer Agent into email, an essential move for mid-market and enterprise support operations where email still carries significant volume. HubSpot’s leadership frames 70% autonomous resolution as a checkpoint, not a ceiling, arguing that as frontier models improve, AI agents will steadily move up the complexity ladder. The key question for contact center leaders is how quickly that trajectory will continue—and how to redesign workflows around it.
RingCentral’s Omnichannel AI Receptionist Expands Its Reach
While HubSpot pushes deeper into ticket resolution, RingCentral is reshaping the front door of customer interactions with its omnichannel AI receptionist. The AI Receptionist (AIR) product now integrates with Shopify, Calendly, and WhatsApp, extending its role beyond basic call answering into routine customer service tasks. Through Shopify, AIR can handle order enquiries over the phone; via Calendly, it can schedule appointments; and WhatsApp support lets it respond to inbound messages on one of the most widely used messaging apps. AIR is also being added to shared SMS inboxes and call queues, stepping in when lines are busy or staff are unavailable. With more than 11,800 businesses already using AIR—particularly smaller and mid-sized organisations—the product is increasingly positioned as a digital front-desk employee that provides always-on support without building a traditional contact center.
Real-World Gains: Shorter Waits and Higher Satisfaction
Early adopters of RingCentral’s AIR are reporting tangible improvements in service metrics that matter to both customers and leadership. Keller Interiors, which deployed AIR across 33 locations, used it to solve a complex routing problem that was difficult to address with human staffing. The company cut average waiting times from 12 minutes to 90 seconds and recorded a three-point rise in customer satisfaction within four months. Maple Federal Credit Union reports reducing hold times by 90%, freeing staff to focus on conversations that truly require human judgment. Features like automatic language detection—supporting ten languages including English, Spanish, French, Italian, German, and Portuguese—help AIR handle a broader range of callers without manual transfers. Analysts describe these updates as a model for applied AI: every feature tied directly to a specific operational pain point, from routing to after-hours coverage.
AI Agents as Core Contact Center Infrastructure
Together, HubSpot and RingCentral illustrate how AI agents are becoming standard infrastructure for contact centers and customer-facing operations. On one side, high autonomous resolution rates allow AI customer support agents to take over tier-one and after-hours workloads. On the other, omnichannel AI receptionists orchestrate calls, messages, and appointments across voice, SMS, and apps like WhatsApp while plugging into tools such as Shopify and Calendly. Integration with existing business systems is the common denominator, enabling seamless deployment without ripping out current workflows. As AI credit consumption rises and products like AIR are packaged as standalone offerings, organisations of all sizes can test value quickly—often through free trials or low-cost entry tiers—before scaling. The emerging pattern is clear: AI agents are no longer optional add-ons but foundational components of how modern customer support is delivered and measured.