Of all the companies I track, Zoom Communications Inc. might be the most interesting, as it’s evolving in ways that run counter to its traditional peers.

When speaking with industry colleagues, including investors, channel partners, customers and fellow industry analysts, Zoom is often grouped into the unified communications-as-a-service or the contact center-as-a-service bucket. Zoom was a pioneer in video meetings and has used its expertise in helping people connect to expand its portfolio in several directions. It is now mainstream across phones, contact centers, messaging and other communications-centric capabilities.

However, recently the company has diverged from its peers and has been looking to disrupt the way people work, including communications, front-line worker tools, documents, e-mail and other forms of collaboration. Earlier this year, at Enterprise Connect, Zoom outlined its vision to become a “system of action” company, designed to reshape work.

The shift has left many industry pundits scratching their heads and wondering whether Zoom can succeed. I do like the shift and believe that, if Zoom is going to achieve its goal of becoming a mainstream work platform, it’s the right move. The best proof of success is customer adoption, as reflected in the most recent first-quarter 2027 earnings results.

From collaboration to completion

At its industry analyst event, Perspectives, in April, Zoom argued that the future of work is “agentic.” Humans talk, artificial intelligence acts and workflows complete themselves across tools without users tab-hopping or re-keying data. Instead of optimizing meetings or chat as standalone apps, Zoom is building a system of action that connects conversations, workflows and outcomes across Zoom Workplace, Phone, Contact Center, Workvivo and AI services.

Chief Executive Eric Yuan’s talk track echoed this on the earnings call. Yuan (pictured) described today’s “private AI era” as still a two-step process. A rep has a Zoom call, then manually updates the customer relationship management system; a doctor talks to a patient, then spends extra time in Epic. In Zoom’s model, that becomes one step: “After the Zoom conversation, my agent will automatically get the work done for me,” with meetings generating a rich context layer that AI uses to drive the next actions.

Zoom’s bet is that human-to-human interaction will become more important, not less, because it is the primary source of high-value context for AI. On the role of people, Yuan was crystal-clear: He does not envision “your agent and my agent talk with each other” as the dominant pattern; instead, humans will continue to collaborate while agents quietly handle follow-through. That is a sharp contrast to many AI narratives that imagine agents displacing human communication.

AI Companion and My Notes as the new UX

The future of work has a very different user experience. AI Companion 3.0 and My Notes are emerging as the front door into agentic work, not just “copilots” bolted onto meetings.

In Q1, paid monthly active users of AI Companion grew 184% year-over-year, driven by early adoption of 3.0 capabilities that extend beyond summaries to include agentic retrieval and workflow orchestration across Zoom and connected applications. Yuan called out My Notes as a breakout product, surpassing 1.5 million monthly active users (excluding trials) just four months after launch, giving users “a personal AI note taker that captures context across Zoom, in-person and third-party meetings” so they can stay present while the system turns conversations into structured takeaways and action items.

The important shift is what happens after the meeting. Yuan explained that with the custom AI Companion, users can attach a workflow when scheduling a Zoom call: During the meeting, My Notes captures context; when the meeting ends, the workflow “automatically takes over to get something done for you.” That is the system-of-action thesis in miniature: Zoom is no longer just the place where work is discussed; it becomes the orchestrator of the work itself.

During the call, Zoom shared customer examples as proof points.

• Raymond James started with AI Companion for summaries, then expanded to a custom AI Companion across about 10,000 seats, enabling wealth advisers to access tailored workflows and summaries that meet their industry’s security and compliance requirements.
• MongoDB upgraded to Zoom Workplace Enterprise Plus, Zoom Contact Center, and Zoom Virtual Agent or ZVA, using a custom AI Companion to translate live conversations into completed actions across information technology ticketing, CRM and other systems.

These are early examples of agentic work at scale: Conversation is the trigger, but the outcome is a ticket filed, a CRM field updated or a workflow executed — without additional human clicks.

A horizontal ‘context layer’ for the enterprise

One of the more nuanced future-of-work ideas on the call was Yuan’s description of Zoom as a cross-stack AI company, not just an app. He acknowledged that customers today mostly view Nvidia Corp., OpenAI Group PBC and Anthropic PBC as “AI companies” because they operate at the chip and model layers, but argued that over time, “application-layer AI companies” will emerge and that Zoom wants to be one of them.

The lever is data. Real-time, multimodal conversation data is an enterprise asset, and Zoom’s position across meetings, phone, contact center and employee experience gives it a uniquely broad dataset. Yuan called Zoom conversations the “context layer” that AI uses to generate insights and tasks; after every call, there is a rich record of intent, sentiment, commitments and entities that can drive downstream automation.

Zoom AI Services is the clearest expression of that idea as infrastructure. Launched in March, it exposes Zoom’s speech recognition models as application programming interfaces. The Scribe model is already ranked among the top automatic speech recognition systems on the Hugging Face open leaderboard, with early adoption from business process outsourcing firms such as InflexionCX Inc. That’s Zoom essentially making a core capability into a product, honed across “countless daily meetings” and turned into a horizontal AI primitive other developers can build on.

The same pattern is evident across verticals. BrightHire, which Zoom acquired, uses conversational AI on top of Zoom to improve recruiting and hiring decisions, winning customers such as Figma Inc. and expanding at HubSpot Inc. with BrightHire Screen, an AI interviewer for go-to-market hiring. Seer by Workvivo extends employee comms into “AI-powered people intelligence,” helping leaders listen to feedback, measure engagement and act through built-in communication tools. In each case, Zoom is less a meeting tool than the platform on which more specialized, AI-driven workflows run.

Rethinking customer experience around AI and conversation

Customer experience is often where future-of-work chatter meets operational reality, and this is where Zoom’s system of action story is furthest along. Zoom positions itself as one of the few scaled vendors with a truly native platform that bridges unified communications and customer experience, connecting collaboration, voice, contact center, virtual agents, expert assist and analytics.

In practice, that means a customer interaction doesn’t end at the IVR; context can carry from an inbound call handled by ZVA to a live agent on Zoom Contact Center to a follow-up meeting, all within a single AI workflow and data model. Yuan contrasted this with CRM-centric approaches, arguing that Zoom is entering the market “from a different angle,” where the contact center is “conversation centric rather than system-of-record centric,” and where the same infrastructure that powers meetings also powers CX.

Q1 showed how that plays out:

• Zoom customer experience grew at a high double-digit rate, with paid AI featured in nine of the top 10 ZCX deals and ZVA included in four of those top ten.
• Chelsea FC chose Zoom Phone, Zoom Contact Center Elite and ZVA Chat to modernize fan engagement, using these platforms to deliver more personalized experiences and build a connected data foundation for long-term growth.
• Caliber Collision deployed Zoom Phone with Zoom Contact Center Elite across more than 1,800 repair centers and a central contact center to eliminate “cold call” experiences and provide unified CX analytics from first touch through resolution.
• Japanese customer Renza selected ZVA, Agentless Dialer and ZCC Elite to modernize high-volume interactions and is using ZVA in a differentiated outbound use case — pre-confirmation calls tied to electricity and gas connections — freeing humans for higher-value sales work.

Yuan highlighted outbound ZVA as an especially large opportunity, noting that enterprises increasingly want to use AI to reach more customers and prospects, and that Zoom is evolving pricing toward “Autocon-based” models that align with outcomes rather than raw minutes. That’s a future-of-work story too — less about staffing more agents, more about orchestrating a blend of automation and human expertise against a unified context graph.

Where the future of work vision is ahead of the market

Despite all the progress, Yuan was candid that Zoom still has an awareness issue. Even some employees “did not realize you can do this, can do that” with the new AI-driven workflows. He acknowledged that customers don’t yet reflexively think of Zoom as an AI company the way they do model providers. He framed that gap as an opportunity, promising “very exciting new product solution announcement next month, all AI-driven” and a renewed go-to-market push to make the system-of-action value proposition explicit.

Chief Financial Officer Michelle Chang offered important commentary, downplaying the value of reporting a standalone “AI revenue” figure and instead saying the key test is whether AI “inflects your total revenue growth” and reduces churn by being embedded in paid SKUs at no additional cost. That framing fits the future-of-work narrative: AI isn’t a feature you upsell; it’s the fabric that makes collaboration, CX and employee experience meaningfully different from what they were in the SaaS 1.0 era.

Zoom’s Q1 shows an organization trying to redesign work around the atomic unit of conversation, with AI as the execution layer and workflows as the outcome — and starting to see real customers run their business that way. In a market still obsessed with seats and suites, that is a genuinely differentiated view of where work is headed.

Zeus Kerravala is a principal analyst at ZK Research, a division of Kerravala Consulting. He wrote this article for SiliconANGLE.

Photo: Zeus Kerravala