Inside Flip: The Vertical AI Voice Agent Powering $12M+ in ARR

What happens when a scrappy college ride-sharing app pivots into AI at exactly the right time? You get Flip: a vertical AI voice agent that now automates hundreds of millions of customer support calls and generates well into eight figures of annual recurring revenue.

Co-founder and CEO Brian Schiff has spent nearly a decade building toward this moment, turning a small Cornell project into one of the fastest-growing voice AI platforms in customer support.

From Campus Ride-Sharing to Voice AI

Flip’s story starts in 2015–2016, when Brian and his co-founder Sam were freshmen at Cornell. Ride-sharing was the hot startup category, but services like Uber were banned in Upstate New York. They spotted a gap and launched a local ride-sharing app across Cornell and other nearby schools.

By 2018, it became clear that the ride-sharing model they were pursuing wasn’t a great long-term business. That forced a critical decision: shut everything down and move on, or pivot into a bigger opportunity.

They chose the pivot. Drawing on their experience with transportation operators and phone-based bookings, they asked a new question: instead of building another consumer app, what if they could optimize all the business already happening over the phone?

That question became the foundation for Flip, an AI voice assistant built to handle customer support calls end-to-end.

What Flip’s AI Voice Agent Actually Does

Flip focuses on one big problem: the flood of customer support calls into large consumer-facing businesses. Rather than generic chatbots or clunky phone menus, Flip offers a natural-sounding AI voice agent that can understand callers, access backend systems, and complete tasks without a human agent.

Today, Flip is deeply specialized in three industries:

Transportation – ground transportation, ride services, and dispatch operations

Retail & E-commerce – brands that handle high volumes of customer orders, returns, and inquiries

Healthcare – especially urgent care clinics with heavy inbound call volume

Instead of being a horizontal, one-size-fits-all AI platform, Flip is intentionally vertical. It builds deep integrations, workflows, and conversation patterns tailored to each industry.

Deep Vertical Focus: Why Flip Isn’t Just “An AI Wrapper”

In AI customer support, there are two main approaches:

1. Horizontal platforms that try to serve any company in any industry with a generic toolkit.

2. Vertical solutions that go deep into a few industries and ship ready-made workflows, integrations, and best practices.

Flip is firmly in the second camp. The team believes that for high-stakes, high-volume phone interactions, depth beats breadth.

Example: Automating Transportation Calls

One of Flip’s flagship transportation customers is A2B Transportation, a conglomerate that coordinates much of the ground travel across Australia. Their contact centers receive tens of thousands of calls per day from people who need to:

• Schedule rides

• Modify or cancel bookings

• Change pickup locations

• Get price quotes

• Connect with drivers

Flip’s AI voice agent automates 85–90% of those calls. It does this by:

• Quickly establishing credibility so callers don’t feel like they’re stuck in a 1990s IVR menu

• Identifying why the person is calling

• Using backend integrations with dispatch and booking systems to actually complete the task (not just answer FAQs)

The result: faster resolutions for customers and dramatically lower load on human agents.

Out-of-the-Box Workflows, Not Custom Builds

One of Flip’s biggest advantages is that new customers don’t have to start from scratch. Because the platform is industry-specific, about 95% of what a new company needs is already built into the product.

That includes:

• Prebuilt integrations with phone systems and dispatch/commerce/clinical platforms

• Standard workflows for common call types (booking, rescheduling, order status, clinic information, etc.)

• Conversation flows battle-tested across hundreds of customers and millions of calls

New customers mainly configure business rules, branding, and edge cases rather than designing an AI system from the ground up. That’s a sharp contrast to some competitors that require large, multi-year, highly customized deployments for each enterprise account.

Pricing, Scale, and Revenue: How Big Is Flip?

Flip uses a usage-based pricing model that roughly aligns with “jobs to be done” on each call. The exact structure varies by industry and call complexity.

In retail, the public rate is around $1.50 per resolved call. In transportation, where calls tend to be shorter and simpler, the rate is a fraction of that. Across all verticals, typical annual contract values (ACVs) range from $50,000 to $500,000 per customer.

Key numbers shared by Brian:

• 250+ customers across transportation, retail, and healthcare

• 300 million+ automated phone calls to date

• At least one customer already paying over $1 million per year

• Well into eight figures of ARR – directionally above $12.5M in annual recurring revenue

• ~75–78% gross margins, indicating there’s substantial proprietary value beyond just calling an LLM API

Flip has been growing roughly 3x year-over-year for the past few years, while keeping burn relatively modest compared to many heavily funded AI support startups.

Funding, Valuation, and Why They Turn Down $150M

Flip’s financial journey mirrors its product evolution:

• 2018 – Pivot from ride-sharing to voice AI, start in transportation

• 2020 – COVID hits; transportation usage drops 80–90% overnight, forcing tough decisions but ultimately survival

• 2021–2022 – Cross ~$1M ARR, expand into retail and e-commerce, land early customers like Brooklinen, then larger brands like Under Armour, Tory Burch, and Authentic Brands Group

• 2021/2022 – Raise an $8.5M seed round to fuel vertical expansion

• 2024–2025 – Add healthcare as a third vertical, focusing on urgent care and serving thousands of clinics

• Early 2026 – Close a $20M Series A at a valuation “around $100M,” with multiple term sheets on the table

During the Series A process, acquisition interest naturally surfaced. According to Brian, offers in the range of $150M all-cash wouldn’t be enough to tempt him and Sam to sell. After nearly a decade of grinding through product-market fit, a major pivot, and a pandemic shock, they feel the most exciting phase of the business is just beginning.

They also deliberately chose “sober, high-quality investors” over the highest valuation term sheet, prioritizing long-term partners over short-term optics.

Why These Verticals: Transportation, Retail, and Healthcare

Flip could theoretically apply its technology to any industry that handles phone calls, so why these three?

Brian points to two key criteria:

1. High-volume consumer interactions. The model works best where lots of everyday people are calling for help: booking rides, asking about orders, or scheduling medical visits.

2. Competitive pressure to innovate. Industries with intense competition are forced to improve customer experience and adopt new tech faster. That’s less true in oligopolies like utilities or some financial services segments, where contact centers are often still not even fully in the cloud.

Transportation, retail/e-commerce, and urgent care healthcare all check both boxes: huge call volumes and strong incentives to deliver a better, faster, more modern experience.

Where Voice AI Fits in the AI Stack

Voice AI is quickly becoming one of the most practical applications of large language models in business. Alongside AI coding assistants, AI customer support is one of the top use cases enterprises are actively deploying today.

Flip sits at the intersection of several trends:

• The shift from generic chatbots to specialized AI agents that can actually take actions

• The move from broad, horizontal AI platforms to vertical solutions with deep domain knowledge

• The growing expectation that customer support should be instant, 24/7, and high quality, without exploding headcount

If you’re interested in other ways AI is transforming voice and audio, it’s worth exploring tools that let you clone or generate voices, like the ones covered in this guide to running a free AI voice generator on your PC. And if you’re thinking beyond single agents toward orchestrated AI systems, you may also want to look at how agent frameworks are evolving, as discussed in this deep dive on making AI agents actually useful.

The Road Ahead for Flip

Flip’s team believes we’re at a moment where AI-powered customer support is no longer a wild idea—it’s “obvious, inevitable, and imminent.” Every CX leader now needs a credible AI strategy, and the bar is rising quickly.

With 250+ customers, hundreds of millions of calls handled, strong margins, and fresh capital, Flip is positioning itself as the default AI voice layer for three massive industries. The company’s next big milestone: pushing past $50M in ARR while maintaining its focus on deep, vertical specialization.

For now, Brian and Sam aren’t interested in cashing out. After years of searching for product-market fit and stabilizing the business, they’re finally in the phase where the machine is working—and they want to see how far it can go.