Adi Patel on Pivoting Out of the Wrapper Era: When Defensibility Beats Distribution
First Founder Dinner for S26: Co-founder and CEO of Clearline. Queens alum. YC S22. Built and killed a 50,000-user LLM product before betting the company on AI for car dealerships.
TL;DR — Adi Patel runs Clearline, an AI-native operating system for car dealerships, with his brother and co-founder Abhi as CTO. They went through YC in Summer 2022 as Lancey, a front-end demo cloner. Lancey pivoted into LLM-powered feedback management, scaled to 50,000 users, and got caught in a token-pricing arms race. Adi shut it down. Clearline is the current bet: voice AI for an industry that has never bought AI before. The dinner covered what it actually looks like to kill something that’s working, why Clearline operates out of a Kitchener office one floor below Builders Club, and how a young non-technical CEO walks into automotive trade shows and walks out with serious contract conversations.
The first Founder Dinner of the S26 cycle. A long table, a full room, and the kind of conversation that only happens when the speaker is close enough to the work to still be in it. Adi didn’t have to travel far. Clearline works out of an office in Kitchener directly below Builders Club, which means most of the room has walked past their door at some point this year. That proximity showed up in the conversation. Several of the questions weren’t hypothetical; they were the questions you ask the team you actually share a building with.
Adi is the co-founder and CEO of a YC S22 company that has pivoted twice. Clearline is where they’ve landed, and it’s in the middle of the customer conversations that will determine whether the bet pays off. He talked about all of it openly, which set the tone for the rest of the evening.
From Lancey to Clearline
Adi and Abhi started Lancey in 2022 as a front-end cloning tool. It was the kind of thing engineers built when GPT-4 made it possible to scrape a website and stand up a working demo in a few minutes. It was useful. It also wasn’t a company.
The first pivot came through customer feedback. Lancey moved into feedback management, using early LLMs to help product teams parse signal from noise across user inputs. The product worked. It scaled to 50,000 users. Revenue was decent. By most early-stage metrics, this was a success story.
It was also, in Adi’s words, undefensible. The core mechanic was token-based pricing on top of a foundation model. Every quarter a new competitor showed up with better prompts and a thinner margin. The product was a wrapper, and the wrapper economy had stopped being a place where good companies got built.
Killing a working product is harder than killing a failing one. The reflex is to keep going because the numbers don’t tell you to stop. Adi stopped anyway.
Why Car Dealerships
Clearline is sharply scoped. The product is an AI-native operating system for car dealerships, built around three pieces: an inbound voice agent that answers every sales and service call, an outbound system that runs disciplined follow-up cadences from day one through day ninety, and a CRM that logs everything so managers actually know what’s happening across sales, service, and the back office.
The numbers in the dealership world are absurd in the way that markets ignored by software usually are. Twenty-three percent of dealership calls go unanswered. Eighty-five percent of those callers never call back. Lead follow-up typically dies by day five. The existing stack is fragmented and nothing talks to each other. Clearline’s early customers are recovering sixty thousand dollars or more in their first thirty days.
The thesis underneath the wedge is broader. Most of the durable AI value over the next decade lives in industries that haven’t seriously encountered modern AI yet. Not SaaS dashboards. Not knowledge-worker copilots. The physical-world economy: service businesses, logistics, retail operations, automotive. The places where the existing tech stack is a phone tree and a clipboard. Voice is the entry point. The CRM is the lock-in.
Selling Into Industries Without AI
The most useful thread from the dinner was how Adi finds and closes customers in an industry that has never bought AI before.
One of the larger contracts now in flight, a 200-plus-location automotive chain, came out of a 400-person aftermarket automotive conference. Adi was the youngest person in the room by a meaningful margin. He used that. People wanted to talk to him because he was curious-looking, not threatening, and because he wasn’t there with a slide deck. He was there asking questions. The conversation started from a hallway introduction and is now three months in.
A few patterns came out of the broader conversation about selling into AI-naive industries. The first is that you can’t lead with the technology, because nobody on the other side has a reference point. You lead with the outcome and let them ask how it works. The second is that ROI alone doesn’t close deals at the operations level. The decision-maker is a regional manager with a bonus structure and a career path. Helping them get promoted matters as much as the cost savings. The third is that the existing AI exposure of your buyer is almost always invisible to them. They use Google Maps, Netflix recommendations, fraud detection on their credit cards. Walking them through what they already trust is more effective than walking them through what you’re building.
A Floor Below Builders Club
Clearline operates from Kitchener on purpose. The reasoning is operational, not sentimental. First-year engineers come in at a fraction of the cost of comparable hires in San Francisco. The talent pool is technically deep. The team is on a five-day in-person policy, communicated up front in job applications, and that filter works in this city in a way it doesn’t elsewhere.
Adi’s co-founder and CTO is his brother Abhi. The company is committed to the location. The office is one floor below the room where this dinner happened. That kind of density, where a YC company and a founder program share a building and most of the founders in the room recognize each other from the elevator, is the thing that makes a place actually compound over time. It’s also the thing that’s hardest to engineer on purpose.
What Adi’s Presence Represented
The first dinner of a series sets the temperature for the rest of it. The room this year is largely founders at the earliest stage. Adi is a few years further along, and he can still describe the early stretch clearly enough that the lessons translate. He talked about killing a product that was working, what it actually looks like to sell into an industry that’s never bought AI, and what it took to get from a YC demo to a company with customers. None of it was abstract.
That’s the version of founder storytelling this series exists for.
Key Takeaways
Working is not the same as defensible. Lancey hit 50,000 users and meaningful revenue, and Adi shut it down anyway. The signal he trusted was structural, not metric: every quarter a competitor showed up with the same product and a thinner margin. A growing wrapper is still a wrapper.
Physical-world AI is an education problem before it’s a technology problem. Most of the addressable market for AI in service industries hasn’t bought AI before. The sales cycle is long because you’re teaching, not pitching. The companies that win this wave will be the ones who learn to sell into curiosity rather than urgency.
Show up where your customers already are, not where founders already are. A 400-person aftermarket automotive conference is not a place where AI startups hang out. That was the point. Adi was the only one in the room talking about what his product did, and he was the youngest person by twenty years. Both of those things were assets.
Sell to the person, not just the company. Operational buyers in mid-market industries have careers, bonuses, and personal stakes. ROI gets you in the room. Helping them get promoted closes the deal. Founders who only optimize for the spreadsheet miss the human variable that decides most contracts.
Density beats geography. Clearline operates one floor below Builders Club in Kitchener. That arrangement isn’t an accident. Cheaper engineering, deeper technical pool, a culture that supports in-person work, and the kind of accidental encounters that turn into customers and team members. Distributed teams are a strategy. Being physically close to other builders is a different one, and it compounds differently.
The bet that matters is the one you can articulate clearly. Adi can say what Clearline is in one sentence: AI for the physical economy, starting with car dealerships. That clarity didn’t come from a positioning workshop. It came from pivoting twice and learning what each version taught him about where the durable value lives.
The S26 Founder Dinner Series is hosted by Builders Club in Waterloo. Sessions pair evening founder dinners with midday workshops. Thanks to our sponsors Osler, TD Innovation Partners, and Communitech for their continued support.