Problem

Darkness makes ordinary visible-light cameras photon-starved, while traditional analog image intensifiers and thermal cameras impose trade-offs in cost, resolution, and image quality. Deepnight targets the operational and economic problem of enabling reliable vision in moonless, starlit conditions without requiring expensive thermal hardware or specialized image-intensifier systems. Its clearest use case is defense: helping soldiers, autonomous systems, drones, and surveillance infrastructure see and navigate in environments where conventional cameras fail.

Product / Service

Deepnight provides software-defined computational night vision. Its deep-learning model processes data from low-light CMOS or other visible-light sensors, aggregates photons over time, and reconstructs meaningful, high-fidelity imagery—even in conditions illuminated only by ambient starlight. The model can run directly on an edge-AI chip in a small camera, reportedly delivering 90 frames per second at one watt, while automatically adapting to motion and changing environments.

The company’s delivery model is primarily software and embedded technology supplied through hardware partners, including camera, goggle, helmet, robotics, drone, and vehicle manufacturers. This approach can turn relatively inexpensive, commoditized electro-optical cameras into night-vision systems, potentially providing higher resolution than thermal cameras at lower cost while avoiding replacement of the customer’s entire hardware platform.

Market

Deepnight competes in defense technology and the broader computational-imaging, digital-night-vision, and low-light-camera markets. Its alternatives include incumbent analog image-intensifier products such as those associated with L3Harris, thermal cameras used for autonomous navigation and surveillance, and digital low-light suppliers such as SIONYX. Rather than selling only a finished goggle, Deepnight is positioning an AI layer as a way to replace or augment conventional night-vision hardware across military, autonomous-vehicle, drone, infrastructure-security, wildlife, and agricultural applications.

The company appears to have moved beyond a pre-revenue technology demonstration, although publicly available evidence does not establish recognized revenue. TechCrunch reported about $4.6 million in federal-government and corporate contracts, including the Army, Air Force, SIONYX, and SRI International, after an initial $100,000 Army contract. Deepnight also raised a $5.5 million round led by Initialized Capital with Y Combinator and other investors, and later partnered with Picogrid and Circle Optics on defense and Air Force-installation applications, indicating early government, industrial, and ecosystem traction.