Flai: The Startup Transforming Car Dealerships with AI Innovation

Silicon Valley is renowned for its comfortable offices filled with polished meeting spaces and on-site coffee specialists. So, when early HappyRobot engineer Ari Polakof set out to start his own business, he certainly didn’t envision himself working on his laptop in a bustling service bay next to mechanics.

“It’s very noisy, impossible to focus,” he laughed.

Nonetheless, this experience is part of the story behind how Polakof’s AI-powered startup for car dealerships, Flai, was born last year.

Co-founded by Polakof and his brother Alen (also from HappyRobot) at Y Combinator, along with former Netflix data scientist Juan Alzugary, Flai is one of several startups looking to harness artificial intelligence to improve the buying, selling, or servicing experience at car dealerships.

The co-founders have created software tailored specifically for the dealership environment. As Polakof explained in an interview, it’s “omni-channel,” capable of handling phone calls (via voice agents) as well as emails and texts (using large language models).

Flai has successfully raised a $4.5 million seed round to expand its operations. Liz Wessel at First Round Capital led the round, with contributions from YC, RedBlue Capital, Joe Montana’s Liquid 2 Ventures, and Innovation Endeavors.

Flai is not alone in this endeavor. Earlier this year, another Y Combinator-backed startup called Toma announced it had secured $17 million from investors like a16z and Yossi Levi, known as the Car Dealership Guy. Numerous others are pursuing similar solutions, while established interactive voice response firms are striving to keep up.

As CEO, Polakof emphasized that Flai stands out by not depending on generic voice technology. Rather, the startup has developed almost everything internally, creating voice agents so convincing that some dealers have already switched from their competitors.

Additionally, he pointed out that there is ample room for competition in the market. Many car dealerships and service centers nationwide are grappling with a common challenge: losing potential customers when their phone lines are busy.

Since launching last year, the Flai team discovered that the most effective way to train their AI and attract customers was through personal visits to those dealerships.

They tried cold calling, emailing, LinkedIn outreach, and even attending conferences. However, most of their early efforts involved directly visiting around 400 physical locations, according to Polakof.

Once they established initial relationships, Polakof noted they fully integrated themselves with each dealer. This led to him spending time in a service bay last year — though more frequently, it simply meant setting up in an empty office.

Polakof shared that the Flai team was on the road “every day, all day” during that critical phase. “It’s almost very tiring, painful, but I believe it’s the only way, and I can’t envision a larger company doing it,” he remarked.

As the startup transitions from this learning stage to focusing on growth and customer satisfaction, Polakof mentioned that the small team continues to put in long hours — much like many of their peers in Silicon Valley.

Although the seed funding will help Flai in its expansion, Polakof indicated he doesn’t expect a significant increase in team size.

“We want to operate as efficiently as possible,” he clarified. “We’re not looking to grow to 100 or 200 people right now.” Instead, Polakof stressed that their three-person team holds each other “accountable” by concentrating on tasks that are “100% necessary; otherwise, it’s not worth it.”

In this sense, Polakof remains quite connected to those early days spent in a service bay or dealership back office — just now with a bit less noise and fewer mechanics around.