Understand any stock in simple English
Enter a ticker to learn what the company does and how it makes money.
DeepGlint sells AI-powered computer vision software and hardware that help cities, stores, banks, and transit systems recognize people, objects, and behaviors in real time.
DeepGlint makes the "eyes" that machines use to understand the physical world. Its core technology is computer vision — software that watches camera feeds and turns raw pixels into useful information: that's a person, that's a car, that's someone jumping a turnstile or leaving a bag behind. Cities, banks, retailers, sports venues, and rail operators buy this so they can monitor crowds, spot safety risks, count customers, or speed people through security without a bottleneck.
The business splits into two main parts. The bigger one is selling AI products and solutions — cameras, edge devices, and software packages tuned for specific jobs like urban management or subway security. The other is its underlying platform, a set of vision algorithms and tools it builds on, which is where the long-term technical edge comes from. Customers pay for the hardware and the software together, often as a project, and they choose DeepGlint because building this kind of vision AI from scratch is hard and slow. Competitors in China's AI vision market include Hikvision and SenseTime, though DeepGlint has carved out a niche in demanding, government-adjacent projects. Here's the honest part: this is a tough, lumpy business. Much of its revenue comes from government and enterprise projects that get budgeted slowly and paid late. It has burned through cash, and turning a profit has been a struggle — the company has posted losses and its stock has been volatile. The whole sector also faces tighter scrutiny over surveillance technology, which is a real cloud over the growth story. So while the technology is genuinely clever, the money-making is not yet proven.
The bet DeepGlint is making is that as cameras get smarter and cheaper, more customers will want vision AI baked into everyday operations — not just megacity surveillance, but stores that count shoppers and banks that verify identities. Whether that bet pays off depends on winning repeatable, profitable deals instead of one-off projects. For now, it's a technically impressive company still searching for a durable way to make money.
Saved answer · last researched 15 Nov 2026