YC Grad Deepnight Nabs $ 5.5 million for AI night vision software that disturbs a multi -billion dollars industry

Deepnight Lucas Young and Thomas Li co -founders were friends from childhood. Both worked as software engineers on Google, when Young decided he wanted to break the code, so talk, in a problem that had contaminated the US military for decades: digital night vision tech.

Most night vision technology is still proportional. Glasses use optical lenses and a chemical process to turn the minimum light at night into pictures, Young told TechCrunch. And cost from $ 13,000 to $ 30,000 per piece by military contractors such as L3harris and Elbit America.

For years, the US Army has been trying to digitize technology, focusing mainly on material. One case: The budget of $ 22 billion for the integrated Visual Augmentation System (IVAS) project, which Anduril was simply undertaken by Microsoft and Hololens Tech.

Young, who has a degree in Calpoly’s computing photo, spent five years of working software on a smartphone camera. He wrote a code that offset the limitations of tiny opening, cheap $ 50 digital cameras used on smartphones. And Li’s background is in AI technology, especially the vision of the computer.

One day, Young read a scientific document from 2018 called Learning to see in the dark Co-author by the well-known scientist Vladlen Koltun, who is currently in Apple. He discussed the use of AI for low light imaging, but at that time, AI on-device chips were not fast enough to support 90 frames per second (fps) necessary to view it in real time.

In 2024, Young realized that the AI ​​accelerators running on chips (Socs) had progressed enough to support 90 FPS. He talked to his friend Li to abandon his job and set up a start -up called Deepnight. And they immediately entered the winter yacht Y Combinator.

Their wows smartphone app

The army was their first obvious client, but they could not just roll in the Pentagon and close a meeting. Thus, Young found an industrial event where people from the US Army Night vision laboratory were present.

He wrote a White Bible that described his idea: Night vision as a software problem. He gave copies to the event, including a Colonel of Army who agreed to read the paper. “It was just a conversation in the hallway. I wasn’t even in business attire. Only a T -shirt,” Young remember.

The colonel liked what he read enough to put the founders in contact with people in the laboratory, typically known as the US Army C5ISR center.

Desperate to show people that their idea would work, the founders built a Night Vision smartphone application. They put the smartphone in a VR set holding a smartphone.

It was an elementary prototype that was impressive enough to lead to their first sale.

“The army gave us a $ 100,000 contract in February 2024, a month to Y Combinator, based on proof of the concept of a demo smartphone and our white and our presentation,” Young said.

Young and Li then had to make their progress in a more formal demo. The couple flew to Washington DC to show a room full of 10 people how their software worked as well as state -of -the -art glasses, Young said. (Here’s a video on YouTube Show their technology.)

The meeting led to more contracts. One year after the start, the start has closed about $ 4.6 million in contracts by the federal government, including the US Army and Air Force, as well as companies such as Sionyx and SRI International.

Deepnight immediately attracted investors. By the end of the YC, he set a round of $ 5.5 million led by initialized chapter, with angels such as Kulveer Taggar, former In-Q-Tel Brian Shin partner and Matthew Bellamy, head of the band Muse. Y Combinator also fell on their standard agreement.

Perhaps the best of all, Koltun, the scientist who wrote the document that inspired the company, also became an Angel investor.

Deepnight offers software and partners with hardware manufacturers such as glasses manufacturers or military helmets or other products.

“Now we can do everything in the world to see in the dark, because it’s just a software program. So it’s the automotive, security, aircraft, nauticals such as boats, electronics, NAV cameras,” Young describes. And because everything is based on a smartphone off-the-shelf $ 50 camera, their technology does not need expensive specialized material.