Zoom and UJET Bet the Contact Center Belongs to AI Now

You call a company. You wait. A voice picks up. For the entire history of that sentence, the voice belonged to a person in a room full of other people wearing headsets. That is the thing Zoom and UJET are quietly rebuilding - and they are approaching it from opposite ends of the software world, with the same wager underneath.

The category has an unlovely name: CCaaS, or Contact Center as a Service. It is the cloud software that routes a customer question across voice, chat, SMS and video and puts it in front of someone - or, increasingly, something - that can answer. It is not glamorous. It is also one of the few places in enterprise software where you can watch the promise of AI collide with a real, annoyed human at 11pm, and measure whether it actually helped.

UJET, founded in 2015 by a former hardware engineer, has a phrase it repeats like a mantra: AI doing the chores, humans building relationships. Zoom, the company that became a verb in 2020, arrived at the contact center a decade after starting as a video tool. Different origins, different sales motions. The same bet: that agents which never sleep will handle the routine, and the humans will handle the part that still needs a human.

The video company that wandered into support

---

Zoom did not set out to build a contact center. It set out to make video calls tolerable, succeeded beyond anyone's plan, and then noticed something obvious in hindsight: every support desk on earth is a room full of people on calls. In 2022 it launched Zoom Contact Center and started selling it to the customers it already had.

That last part is the whole strategy. Zoom's growth in the category has come less from inventing a better mousetrap and more from adjacency - bundling the contact center alongside the meetings and phone service a company was already paying for. By the second quarter of 2025, Zoom reported more than 1,100 contact center customers, a figure that had roughly doubled year over year. The customers paying six figures grew faster still: 229 accounts with more than $100,000 in annual recurring revenue, up 94% in a year.

The AI piece is where Zoom is spending its attention. In June 2025 it rolled out what it calls agentic upgrades to its customer experience platform. The headline feature: a virtual agent that no longer just answers a scripted question but completes the task behind it - processing a return, updating an account, booking an appointment, across both chat and voice. Around it sit the unglamorous supporting acts - automated scheduling, quality scoring that reviews up to 100% of interactions, topic detection that flags trends without a human reading transcripts. Paid users of Zoom's AI Companion, the assistant threaded through its products, grew 184% year over year.

The specialist that treated support like a product

---

UJET's founder, Anand Janefalkar, came to the contact center from an odd direction. Before starting the company in 2015, he built hardware at Jawbone, HP Palm and Thalmic Labs. He looked at customer support and saw an engineering problem, not a call-volume problem - something to be redesigned rather than staffed around. UJET went mobile-first and AI-first from the start, tightly wired into the CRM systems where customer data actually lives.

Where Zoom sells breadth, UJET sells depth in one thing. That focus has kept it near the top of customer-satisfaction rankings - six consecutive years at number one on G2 for contact center software, and a 2026 ranking as the top CCaaS platform for the mid-market from Info-Tech. It has raised roughly $231 million across five rounds, most recently a $76 million infusion that valued it near half a billion dollars and came with a leadership change: in 2024 Janefalkar shifted to lead product and engineering while former COO Vasili Triant stepped up as co-CEO to run the business.

The clearest signal of where UJET is heading came when it acquired Spiral, a conversational-analytics company built to read support interactions at scale. It is the least flashy kind of acquisition and the most telling. Agentic AI is only as good as its understanding of what customers are actually saying, and buying the data layer that feeds it is how you make the autonomous agent stop guessing. UJET folded it in under the name Spiral by UJET and pointed it at the same goal as everyone else: fewer chores for the humans.

Two playbooks, one thesis

---

It is tempting to frame this as a fight, and in the accounts they both chase, it is. But the more interesting thing is how differently they are betting on the same idea. One sells the whole platform and lets the contact center ride along. The other sells a single deep thing and wins on satisfaction. Both are growing. Distribution and focus turn out to be two valid answers to the same market.

Neither of them is the market leader, and that context matters. The incumbent at the top of the analyst charts has held its position for more than a decade. Zoom's answer has been to undercut it - pricing its suite well below the leader's full offering - while UJET's answer has been to route around it into the mid-market, where a lighter, cheaper, AI-native product can win on fit. Incumbency is a moat right up until the thing it protects stops being the thing customers want.

The question nobody puts on the slide

---

Here is the uncomfortable math sitting under all of it. If a virtual agent resolves the return, updates the account and books the appointment on its own, what is left for the person with the headset? Both companies give the same answer - the hard part, the relationship, the escalation that needs judgment. It is a reasonable answer. It is also the answer every automation wave has given, and customers, not slide decks, get to decide whether it holds.

What makes the contact center a good place to watch this play out is that it cannot hide. Product marketing can oversell; a support queue cannot. The wait time is the wait time. The resolution either happened or it did not. If agentic AI is going to prove it can do useful work at scale rather than demo well, this is the arena where the scoreboard is honest. And the more genuinely interesting question is not whether an AI will answer your next call. It is whether you will be able to tell.

For now, both companies are betting you mostly will not care, as long as the chore gets done and a human is still there when it counts. Whether that bet pays off is the story the next few quarters will tell - one resolved ticket at a time.

Questions people actually ask

---