Tel Aviv-based startup Lightrun, which helps developers debug production code within their IDEs, announced Wednesday the launch of its first AI-powered tool, the Runtime Autonomous AI Debugger. The new tool, currently in private beta, aims to help developers fix issues in production code in minutes, not hours.
Lightrun also disclosed on Wednesday an $18 million SAFE round it raised last year from GTM Capital, with participation from existing investors Insight Partners and Glilot Capital. This brings Lightrun’s total funding to date to $45 million. The company is understood to be planning to raise a Series B round next year.
“So far, we’ve gotten (Mean Time to Recovery) down to an average of 30 minutes, maybe 45 minutes, depending on how we measure ourselves and how we get customer feedback,” Lightrun CEO and co-founder Ilan Peleg told me. “Now, we’re going to automate everything from the moment a ticket is raised to getting down to the root cause down to a single level of detail, like which single line of code is responsible for this very specific root cause.”
Over time, Lightrun would like to expand this to using generative AI to automatically fix bugs, Peleg said. But that’s not an option right now, and given how quickly the technology is advancing, it’s only a matter of time.
To that end, Lightrun is refining its existing model to focus on debugging, something the company can do in part because it gains insights from its entire monitoring and observation stack, not just the code itself. In the future, the company plans to connect the system to other enterprise inputs, such as ticketing systems. “There’s so much data in the enterprise that’s relevant to troubleshooting or debugging that a solution like Copilot doesn’t have,” Peleg said. He argued that most chat interfaces like Copilot don’t have enough contextual insight to suggest the best solution by just looking at the code.
As Peleg notes, the team went through several iterations before they felt the system was ready for everyday use. About half a year ago, Lightrun began experimenting with existing models to see how generative AI could help users. But at the time, the solution was too expensive to ship as a product. “Now we’ve tweaked the system so that it doesn’t add significant cost to the solution. That’s why we’re talking about it now. Before, it was hard to release something that didn’t exist yet.”
For now, at least, these generative AI capabilities will simply be part of the existing Lightrun solution for private beta users. Peleg emphasized that the company is trying to prove that the system actually provides value to users and is not trying to optimize for short-term monetization.
