CarGurus' Approach to AI

Sarah Rich helms the AI Forward team. “There is a lot of nuance within each department in terms of their level of readiness and interest, and the point of AI Forward is to make sure that’s being addressed in every function,” she said. While the mandate for AI Forward is top-down, what it looks like in every function is very bottom-up, Rich said. To that end, AI Forward is intended to meet the unique needs of each department, which require different approaches to experimentation. On the engineering team, this might look like a weeklong “bakeoff” in which engineers get to test out two different AI coding tools to see which works best for them. But another team, like human resources, might take a totally different approach.

According to Director of Engineering Ryan Aghdam, AI Forward is just one example of CarGurus’ culture of continuous learning. “The investment in learning is key in that it goes beyond just learning about tools,” he said. “It’s about tinkering with new ideas and being innovative that way.” In fact, AI Forward has pre-vetted a robust collection of AI tools, which enables people to dive into experimenting with approved tools instead of working to identify and obtain them. As AI makes it easier for CarGurus’ employees to drive innovation, it’s also enhancing its platform. A new AI-powered search experience is a testament to this transformation.

AI has enhanced work for every team at CarGurus. For Makofsky, the AI platform Glean makes it easier for him and his peers in product to quickly search internal data sources, while AI agents summarize findings and automate simple but time-consuming tasks within the platform. Meanwhile, Google’s AI assistant, Gemini, enables his team to easily transcribe meetings and supports their product prototyping efforts. “All of these tools save people time every single day, while giving the whole company more capabilities than we could have imagined, even a year ago,” Makofsky said.