News

Startups

Artificial Intelligence

Americas

Conntour Raises $7 Million to Transform Security Cameras Into Natural Language Search Engines

Conntour has secured $7 million in seed funding to launch an AI platform that allows security teams to search camera feeds using natural language queries.

Conntour has secured $7 million in seed funding to launch an AI platform that allows security teams to search camera feeds using natural language queries.

NewDecoded

Published Mar 27, 2026

Mar 27, 2026

4 min read

Image by Conntour

New Era of Video Intelligence

Conntour, an AI video intelligence startup based in Miami and Tel Aviv, announced today that it has raised $7 million in seed funding to scale its natural language search platform for security cameras. General Catalyst led the round with participation from Y Combinator, SV Angel, and Liquid 2 Ventures. The company aims to replace rigid, predefined surveillance rules with a flexible system that allows operators to query video feeds as easily as a web search.

The platform utilizes proprietary computer vision algorithms to understand context-rich queries without requiring preset categories. Users can search live or recorded footage for specific descriptions, such as a person wearing a specific brand of shirt or a vehicle with unique identifiers. This technology overlays existing camera infrastructure, meaning organizations do not need to purchase new hardware to upgrade their intelligence capabilities.

Founded by computer vision experts Matan Goldner and Tomer Kola, Conntour draws on their extensive experience within intelligence agencies and the high-tech sector. The startup was part of the inaugural Palantir Startup Fellowship and the Winter 2025 batch of Y Combinator. Their approach was born from real-world operational challenges where traditional video analytics failed to provide actionable insights during high-stakes incidents.

According to company data, the platform can reduce manual video review time by up to 90 percent. A single operator can monitor thousands of cameras simultaneously, a feat that typically requires hundreds of personnel. By identifying nuances that traditional motion-detection systems miss, Conntour claims to reduce false alarms by 70 percent while significantly decreasing the rate of missed security events.

The technology is already being utilized in homeland security operations, including paying deployments with government agencies in Singapore. It is designed for high-stakes environments such as critical infrastructure, border security, and large-scale public venues. To meet the privacy and security requirements of these sectors, the platform offers a full on-premises deployment option that keeps sensitive data within the organization's own servers.

Yuri Sagalov, Managing Director at General Catalyst, noted that the founders' first-hand understanding of security operations sets the company apart. He emphasized that the speed at which the team earned trust from government partners highlights the urgent demand for smarter video intelligence. The seed funding will be used to further develop their vision-language models and expand their global footprint.


Decoded Take

Decoded Take

Decoded Take

The launch of Conntour signifies a fundamental shift in the $8 billion AI surveillance market, moving away from simple motion detection toward deep contextual understanding. By applying vision-language models to existing camera networks, the company effectively turns massive amounts of visual data into a searchable database. This evolution mirrors how search engines transformed the early internet, making vast information accessible through simple queries rather than rigid categories. While this technology promises massive efficiency gains for law enforcement and infrastructure protection, it also sets the stage for a new era of intelligence that will undoubtedly challenge existing privacy frameworks and ethical standards worldwide.

Share this article

Related Articles