Resolve AI Raises $40M Extension at $1.5B Valuation to Automate Production Incident Management

At 3 a.m., the pager doesn’t ring, it accuses. It’s not just noise, it’s a system somewhere unraveling while a human tries to stitch it back together with half the story and twice the pressure. Software in production has a way of humbling even the best teams. Spiros Xanthos lived that reality inside observability at Splunk, watching engineers burn cycles and patience chasing signals across fragmented systems. So Spiros Xanthos and Mayank Agarwal didn’t build another dashboard. They built something that walks into the fire with you and knows where to look.

Resolve AI just pulled in $40M in a Series A Extension at a $1.5B valuation, led by DST Global and Salesforce Ventures. That’s not a polite nod from the market, that’s a firm handshake with intent. Especially when you stack it on top of $125M from Lightspeed Venture Partners earlier this year and a $35M seed led by Greylock Partners. Over $190M raised in about 18 months since stepping out of stealth. Some companies ease into the conversation. Resolve AI walked in mid-sentence and finished it.

Now here’s where it gets interesting. While the world’s been busy teaching AI how to write code, Resolve AI is teaching it how to babysit that code in production. Logs, metrics, traces, infrastructure events, all the messy signals engineers usually chase at ungodly hours. Resolve AI doesn’t just watch, it investigates, correlates, and either fixes the issue or hands you the answer like it’s been there the whole time. Coinbase saw investigation time drop by 72%. DoorDash shaved incidents from 40 minutes to about 1 minute. That’s not optimization, that’s reclaiming time people didn’t think they’d get back.

And then there’s Dhruv Mahajan stepping in as Chief AI Scientist to lead Resolve AI Labs. After years shaping large scale models at Meta, Dhruv Mahajan is now focused on something far less theoretical and a lot more unforgiving. Production environments don’t care about demos. They demand accuracy, context, and decisions that hold up when the stakes are real. Resolve AI Labs is betting that general models aren’t enough, and that domain specific intelligence is where the real leverage lives.

What stands out isn’t just the capital or the customer logos like Salesforce, Zscaler, and MSCI. It’s the pattern. Spiros Xanthos and Mayank Agarwal have been here before with Omnition, building in the trenches, then scaling with purpose. They understand that the real constraint in modern software isn’t writing it, it’s running it without chaos.

The takeaway writes itself if you’re paying attention. The winners in AI won’t just generate more output, they’ll reduce friction where it actually hurts. Resolve AI isn’t chasing the spotlight. It’s standing in the dark corners of production systems, quietly resolving what everyone else is still trying to detect. And that’s a different kind of power.