Conntour secures $7M from General Catalyst and YC for AI-powered security video search

The surveillance technology industry is currently under scrutiny, though not for the most favorable reasons. Controversies have flared as U.S. Immigration and Customs Enforcement reportedly accessed Flock’s camera network for surveillance, and home camera maker Ring faced criticism for developing features that allow law enforcement to request neighborhood footage from homeowners. These developments have sparked a broad debate around safety, privacy, and the ethics of monitoring.

Yet controversy does not diminish market demand. Continued advances in vision-language models are fueling growth for companies that help businesses oversee their premises more effectively.

Matan Goldner, co‑founder and CEO of video surveillance startup Conntour, acknowledges that ethical considerations are serious enough that his company is highly selective about its clients. While that might seem counterintuitive for a startup barely two years old, Goldner says Conntour can afford to be choosy because it already counts several large government and publicly listed entities among its customers, including Singapore’s Central Narcotics Bureau.

“The fact that we have such big customers allows us to select them and to stay in control. We’re really in control of who is using it, what is the use case, and we can select what we think is moral and, of course, legal. We use all our judgment, and we make decisions based on specific customers that we’re okay to work with because we know how they will use it,” Goldner told TechCrunch in an exclusive interview.

That traction has done more than just allow Conntour to be selective. Investors have taken notice: the startup recently raised a $7 million seed round from General Catalyst, Y Combinator, SV Angel, and Liquid 2 Ventures.

Goldner said the round closed within 72 hours. “I think I scheduled around 90 meetings in like eight days, and just after three days — we started on Monday and by Wednesday afternoon, we were done,” he said.

Conntour’s pickiness may well be justified, especially given the power of modern AI tools in this field. The company’s video platform uses AI models to let security personnel query camera feeds in natural language, finding any object, person, or situation in real time — essentially a Google‑like search engine tailored to security video. It can also monitor for threats autonomously based on preset rules and surface automatic alerts.

Unlike legacy systems that rely on rigid definitions or parameters to detect specific objects, motion patterns, or behaviors, Conntour says its system uses natural‑language and vision‑language models, offering high flexibility and usability. A user might ask, “Find instances of someone in sneakers passing a bag in the lobby,” and Conntour’s system quickly scans all recorded footage or live feeds to return relevant results.

A screenshot of Conntour’s platform in action. Image Credits: Conntour

Because the platform integrates AI models, users can simply ask questions about footage and receive answers in text, accompanied by the relevant video clips, as well as generate incident reports.

The company’s key selling point, however, is its scalability. Goldner explained that the platform differs from other AI video search services because it’s designed to efficiently scale to systems with thousands of camera feeds. In fact, he said, Conntour’s system can monitor up to 50 camera feeds on a single consumer GPU like Nvidia’s RTX 4090.

The company achieves this by employing multiple models and logic systems, then determining which combination requires the least computing power to produce the best results for each user query.

Conntour claims its system can be deployed entirely on‑premises, fully in the cloud, or a mix of both. It can integrate with most existing security systems or serve as a standalone full surveillance platform.

Still, the video surveillance industry has long faced a persistent challenge: the quality of surveillance is only as good as the footage captured. It’s difficult to discern details from a poorly lit parking lot recorded by a low‑resolution camera with a dirty lens, for example.

Goldner says Conntour hedges against this by providing a confidence score alongside its search results. If the source camera feed lacks sufficient quality, the system returns results with low confidence levels.

Looking ahead, Goldner says the biggest technical challenge is bringing the full capability of large language models to the system while maintaining efficiency.

“We have two things that we want to do at the same time, and they contradict each other. On one hand, we want to provide full natural‑language flexibility, LLM‑style, so you can ask anything. On the other hand, there’s efficiency — we want it to use very few resources, because processing thousands of feeds is just insane. This contradiction is the biggest technical barrier and problem in our field, and it’s what we’re working really hard to solve.”