MayimFlow wants to stop data center leaks before they happen


At mania around the data center there are many companies looking to become “picks and shovels” providers, by building profitable businesses that are ancillary to gold rush gigs selling server access or training leading AI models.

MayimFlow, this year’s World Built stage winner TechCrunch is annoyingis a good example. The startup is focused on one task: preventing destructive water leaks.

Data centers use a lot of water, and that water can present a big risk, even if it’s a small leak. Founder John Khazraee told TechCrunch that many data centers only have reactive solutions to water leaks. That can saddle the company with downtime and set back millions of dollars if anything.

Khazraee will know. He spent more than 15 years building infrastructure for IBM, Oracle, and Microsoft. With MayimFlow, they are developing a combination of IoT sensors and edge-mounted machine learning models that can detect signs of an impending leak.

“I have noticed this problem in the data center, and the only solution they have: ‘when the leak happens, we know,'” he said in an interview. “Now you have to spend a lot of money to fix the situation. Now you have to shut down the server. Now the data is disturbed. So I decided to do something.”

Khazraee assembled a small team with a lot of experience to take on this challenge. Jim Wong, MayimFlow’s chief strategist, has worked in data centers for decades. Chief technology officer Ray Lok has carved out a career in water management and IoT infrastructure.

Beyond the most terrifying consequences of an unknown water leak, there is an element of thrift about what Khazraee is trying to do with MayimFlow, which he says has been going on since childhood.

Techcrunch event

San Francisco
|
13-15 October 2026

“I grew up in a very, you know, I don’t want to say the word ‘poor’ family, but we weren’t the richest family,” she said. “And my dad would always, when I was in the shower, say, ‘Hey, you’ve been in the shower for a long time. Are you singing in there?'”

When he was growing up, Khazraee said he was always thinking about how to make things more efficient – especially when he was studying to be an engineer. In college, he worked at a facility that collected cooking oil from restaurants and turned it into biodiesel. It was a “sneaky project,” he said, but he was happy with the end result.

Khazraee is now trying to blend this penchant for frugality with team experience. They believe they can give data center operators 24 to 48 hours of advanced warning that repairs will be needed.

MayimFlow has collected a large amount of sample data from various industrial water systems that allows them to make such predictions, he said. Companies can provide sensors that can help monitor water systems or plug machine learning models into existing models if the company already has similar hardware.

Khazraee says he wants to bring this solution beyond data centers to commercial buildings, hospitals, manufacturing facilities, and possibly utilities. In his eyes, any company that wants to detect leaks early, or optimize water use, can become a customer.

Khazraee thinks this is a big opportunity as he turned down roles at several large tech companies while building MayimFlow over the past two years.

“I really believe in the vision. I believe in the impact that we make,” he said. “Water is one of the world’s great problems.”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *