The support team of the future doesn't answer tickets. It runs the agent.

Whenever I talk with support leaders about AI, there's usually a question sitting underneath the one they actually ask: what happens to my team? It's fair, and I won't pretend the honest answer is entirely comfortable. In some cases, yes, headcount comes down.

But when I watch how the best support organizations are actually adopting AI, they aren't shrinking. They're changing shape, and in a real sense, getting promoted.

What the airline industry can teach us about support

Here's the analogy I keep coming back to. When the price of jet fuel dropped, airlines didn't simply pocket the savings. They opened more routes. They flew to more destinations. They offered more service. Cheaper inputs didn't shrink the industry. They expanded what it could offer.

Support works the same way. For as long as I've been in this space, running a support org has meant managing a triangle: cost, speed, and quality. You could optimize for two, and the third paid the price. There was only ever so much you could reasonably ask of a team of people.

AI breaks that triangle. Suddenly you can deliver better support at the same cost, or less. And the best teams don't bank the savings and declare victory. They reinvest them into a better experience, the same move the airlines made.

That reinvestment is where the job starts to change.

From answering tickets to running the agent

A support professional's contribution used to be measured in tickets closed. Now it's measured in how well the agent they own performs. They become, in effect, the PM of the agent: defining how it behaves, reviewing where it struggles, deciding what it should be able to do next. They stop working the queue and start running a product.

And let's be honest about something: your customer doesn't care whether an AI or a human solved their problem. They care that it got solved: quickly, accurately, in their language, at whatever hour they needed it. In a lot of those moments, an agent is genuinely well suited to the job. It has infinite patience. It'll work a problem step by step and never give the bare-minimum answer just to clear the queue. Not because people don't care, but because people have a next ticket, a lunch break, and a finite number of languages they speak.

That's exactly what frees your team for the moments that actually need a human. The escalation. The sensitive conversation. The customer who's one bad interaction away from leaving. That's when you want a person in the loop, not a script. AI takes the volume; people take the moments that carry weight.

The shift from reactive to proactive

The biggest change isn't doing the same support more cheaply. It's doing support you simply couldn't do before.

For the entire history of the discipline, support has been reactive: something breaks, the customer reaches out, you respond. Being proactive was a luxury almost nobody could afford. Saying "I noticed you ran into trouble with this last week; here's how to sort it out" would have meant a human monitoring every customer, all the time.

Now it's possible. The agent watches for the moment and reaches out before the customer ever has to. That's where support stops being a cost center you're trying to contain and becomes part of the product experience itself.

What this actually means for your team

So the support team of the future spends less time answering the same question for the thousandth time, and more time managing and monitoring the agent, handling the escalations that matter, and designing an end-to-end customer journey that used to be out of reach.

But here's the part that matters most to me: none of it works if you don't own the agent. You can't be the PM of something you can't see inside, can't change, and don't control. If your agent is a black box someone else operates, your team isn't running it. They're filing tickets against it and waiting. The entire promotion I'm describing depends on ownership: the ability to see every conversation, understand every decision, and improve it yourself, the same day.

That's the future we're building toward at Voiceflow: not AI that replaces your team, but AI your team owns and improves. The companies that get this right won't end up with smaller support organizations. They'll have more capable ones, finally pointed at the work that was always worth doing and never had the time for.