Sparq wants drivers to be their own AI-powered mechanics

Cars are about to get a lot more expensive. This startup wants to make your services cheaper. 

Sparq Diagnostics, cofounded by Codrin Cobzaru and Daniel Nieh, has developed a plug-in scanner designed for everyday drivers. In a market full of specialized tech and pricey mechanics, Sparq aims to give car owners affordable insights into their vehicle’s health. Since launching just three months ago, Sparq has already sold 15,000 devices across Southern California, and is aiming for a nationwide expansion by the end of the year.

“My wife came in crying that she had been taken advantage of by a mechanic that she’d been going to for the past five years,” Cobzaru says. “That’s the reason we started Sparq. The educational component is where we believe there’s a lot of leverage we can bring to everyday drivers. It all starts with knowing what’s going on with your vehicle.”

Inside Sparq’s car tech

Nieh and Cobzaru acknowledge the dozens of OBD-II products (which scan for a vehicle’s Diagnostic Trouble Codes) already on the market—and indeed went through “hundreds” of them during their product-testing phase. But these devices are often built for experienced auto mechanics or scrappy DIYers, not the average car owner. Meanwhile, more consumer-friendly scanners from companies like Nonda and BlueDriver are “not comprehensive enough to tell you what is wrong with your vehicle down to the OEM level,” Nieh says. 

That’s where Sparq comes in. The device translates the insights of an OBD-II scanner into a user-friendly app, presenting information through digestible tools and metrics including a red/yellow/green health indicator and a top-line vehicle score out of 100. And Sparq’s AI chatbot allows users to ask questions about their vehicle and receive personalized answers. Nieh calls this “letting your car speak back to you directly about what it needs.” 

Nieh and Cobzaru founded Sparq in 2021—about a year before ChatGPT’s public debut kicked off the world’s ongoing AI mania. Working with different large language models excited the duo, especially considering the volume of data they were holding onto. AI could make some of these complex car troubles even more digestible. While Cobzaru declined to disclose exactly which models Sparq uses (he notes that there are multiple, calling it their “secret sauce”), he’s quick to separate Sparq from the dozens of specialized AI chatbots. “We had the app built, as well as the device,” he says. “It’s not just a wrapper on top of ChatGPT.”

Plugging the device into my own rental car (a Jeep Grand Cherokee), the diagnostics app quickly picked up on the make and fuel status and, after rapidly performing dozens of tests, rated it  88/100. It flagged a high-voltage condition in my Intake Air Temperature sensor and explained that the defect could cause decreased engine performance and overheating. I asked the chatbot when I should bring the car in for service, and it advised waiting until I hit 48,000 miles.

Fixes like these can be expensive. The average American spends $1,452 for every 15,000-mile year, per AAA. Moreover, Americans are frustrated by negotiating with mechanics. In a survey of 1,000 car owners by American Trucks, 50% of respondents said that an auto repair shop had tried to sell them unnecessary fixes, and 35% said they had been overcharged. 

Interestingly, Sparq’s chatbot estimated the cost of that service at between $275 and $375. This is a key point of leverage: Letting drivers know exactly what a mechanical procedure should cost can empower their negotiations. Cobzaru says their AI generates these labor-cost estimates based on the user’s region and local averages. 

Of course, mechanics aren’t likely to be happy, since Sparq’s labor-cost estimations could cut into their margins. But Cobzaru claims many in the auto service industry have responded positively. “They love when people come in educated because they’re more likely to proceed with a specific service,” he says.

Tracking SPARQ’s expansion

Nieh and Cobzaru had two startups before Sparq Diagnostics. Their first was Shair, an apparel company that Nieh describes as a “brutal learning lesson” for the then-17-year-olds. After that, they founded a gun-safety company called Truss Technologies, which the duo put “on pause” to focus on their latest venture. 

After some initial product development, Nieh gauged consumer interest by running pre-product marketing campaigns. Those efforts netted the founders $51,370 on Indiegogo and $38,417 on Kickstarter. Since then, the company has continued to raise capital, though the two declined to disclose how much they’ve raised to date, citing an ongoing investment round. 

Having built the core Sparq device, the duo is now focused on the future. In the near term, Cobzaru is training Sparq’s AI on the collected-user data, with the goal of reaching “the next level of personalization.” They just announced sound and image recognition in their AI system: If a user scans their tire, Sparq’s AI can estimate their tire pressure. Soon, they’re hoping to have tests for mechanical wear-and-tear built into the device, too, which could go a long way toward crash prevention. 

“Based on how frequently you brake and how hard you break, as well as your speed deceleration, we can know how thick your brake pad is,” Nieh says. “That’s something that we’re going into.” 

Their long-term vision is even broader: Sparq owned-and-operated service centers, a mechanic recommendation system based on user data and more. Says Nieh, “We’re just scratching the surface.”

https://www.fastcompany.com/91313254/sparq-wants-drivers-to-be-their-own-ai-powered-mechanics?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss