Carvana cut its customer care costs by 40 percent in one year.

Then 30 percent the next.

Then 20.

Then 10 this year, four years running, according to CEO Ernie Garcia on the company's second-quarter earnings call.

The tool behind it is an AI customer experience agent the company calls Sebastian.

The used-car retailer's second quarter told the same story from every angle.

Retail units sold rose 38 percent year over year to 197,000.

Net income hit $513 million.

Adjusted EBITDA reached $769 million.

Garcia was direct about the cause: "the progress that we're seeing across the entire business enabled by AI, whether it's just every product in the business being able to move more quickly or if it's customer experiences getting better and simpler, I think that's exceedingly clear."

What stands out is the shape of the savings, not just the size.

Most companies announce an AI win once, in a single quarter, then move on to the next headline.

Carvana has posted a declining customer service cost line for four consecutive years.

That is not a pilot.

That is a system that keeps compounding, because each year's baseline already reflects the previous year's automation.

Garcia made a point that gets lost in most AI coverage: not every gain came from new technology.

Some came from process improvements made before the tooling was in place.

The technology mattered, but so did fixing the underlying workflow rather than bolting AI onto a broken one.

Carvana is large and well-funded, and the scale here is not what a small business will see.

But the mechanism is available to anyone: a single customer-facing agent, assigned to one function, measured and improved every quarter.

The compounding comes from consistency, not size.

For a small business owner answering the same customer questions every day, that is the useful takeaway.

The savings do not arrive from a single dramatic rollout.

They arrive from picking one repetitive job, automating it properly, and returning to improve it on a schedule.

Viktor is built around that same idea of steady, compounding automation, sized for businesses well below Carvana's scale.

Viktor lives inside Slack and Microsoft Teams.

You @mention it in a thread the same way you would ask a colleague.

The output, a PDF, a report, a task created in your CRM, an email drafted in Gmail, lands where it should land.

A small retailer could hand Viktor the recurring customer questions about order status or returns, the same category Carvana targeted first, and get a drafted response ready for review in seconds.

A service business could use it to track which client questions repeat most often, so the owner knows exactly where the next automation should go.

A solo founder could ask Viktor to summarize a week of support tickets into a short report, the same visibility a company like Carvana reports to its own board every quarter.

Viktor is built to SOC 2 and CASA Tier 3 standards, with GDPR and CCPA compliance built in.

Your data sits in an encrypted vault and is never used to train underlying models.

Every meaningful action runs through a Slack approval step before it happens, so nothing moves without a human saying yes.

Full detail is at viktor.com/security.

You get $100 of free credits to begin. No time limit, no commitment. That's enough to do real work and see what Viktor can actually do before you spend a penny. There's also $50 off your first bill.

You must use this exact link to receive both benefits: https://AIThatDelivers.com

Carvana did not fix its customer service costs in one quarter.

It fixed them in four, one year at a time, by returning to the same problem and asking what else could be automated.

That is a more useful lesson for a small business than the size of the number itself.

Disclosure: Some links in this article are affiliate links. If you choose to get started with Viktor using the links provided, I may receive a commission, at no additional cost to you. I only recommend tools I use and believe in.