Conntour raises $7M from General Catalyst and YC for AI-powered video surveillance search

Conntour has raised $7 million from General Catalyst and Y Combinator to build an AI search engine for video surveillance. The startup offers security teams a fundamentally new tool: instead of manually scrubbing through hours of footage, an operator enters a plain-language text query, and the system instantly finds the relevant clips across feeds from hundreds of cameras.

Picture a large shopping mall with a thousand cameras and an incident at the north entrance around three in the afternoon. Previously, an operator would spend hours rewinding footage from the relevant cameras within the relevant time window. With Conntour, it is enough to type "man in a blue jacket, north entrance, 14:30–15:00", and the AI returns the exact clips in seconds.

The technology also handles more complex queries, for example "every instance of a visitor leaving luggage unattended for more than five minutes" or "a person who appeared at the staff entrance several times within an hour". It is built on multimodal language models trained to understand the content of video frames and match it against arbitrary text queries.

The $7 million round was led by General Catalyst, one of the largest US venture funds, whose portfolio includes Airbnb, Snap, Stripe and Hubspot. Y Combinator, the accelerator that launched Dropbox, Airbnb, Reddit and thousands of other companies, also joined the round. YC traditionally invests at the earliest stage; its participation means Conntour passed a tough selection process and gained access to one of the most influential contact networks in Silicon Valley.

The market the startup is targeting is huge. Analysts estimate the global video surveillance market will exceed $80 billion by 2028. Cameras are everywhere: shopping malls, airports, factories, office buildings, residential complexes.

The industry's main problem is not a lack of data but too much of it. A typical large facility generates terabytes of video every day. Finding a specific incident by hand is enormously labor-intensive. AI search solves this at the root: a person describes what they are looking for, and the system finds it in seconds.

The competitive field is already crowded, but the segment is growing fast. Verkada, Avigilon (a Motorola Solutions division) and Milestone Systems are all integrating AI features into their platforms. Among specialized startups, Spot AI, BriefCam and Arcules stand out. Conntour's difference is that it makes natural language the primary interface rather than a supporting one. That substantially lowers the barrier to entry for non-technical users, such as guards, facility managers and compliance specialists, who do not need to learn complex system settings.

The funds will go toward expanding the engineering team, refining the product and landing the first large enterprise contracts.

Conntour is a typical example of a broader shift: AI is turning surveillance systems from passive archives kept "just in case" into active analysis tools. Camera recordings stop being mere storage and become a full-fledged database that can be queried the way you query spreadsheets or enterprise search. If the startup delivers on its promise at the scale of real enterprise deployments, it would change security operations standards worldwide.