Matan Goldner still sounds a little surprised by how quickly it happened. Eight days of meetings—ninety, by his count—compressed into a blur of Zoom calls and term sheet negotiations. By Wednesday afternoon, it was over. Conntour, the startup he co-founded with Tomer Kulla to make security camera footage searchable through plain English, had closed a $7 million seed round.
The speed matters. Seventy-two hours from verbal commitments to signed documents is unusually brisk, even in an ecosystem conditioned to move fast. It suggests something beyond typical venture enthusiasm—perhaps a recognition that infrastructure plays in defense-adjacent markets, particularly those handling sensitive government data, have become investable in ways they weren't just a few years ago.
General Catalyst led the round, announced in late March, with Y Combinator, SV Angel, and Liquid 2 Ventures joining. For a Miami-based company founded just two years ago, it's the kind of validation that opens doors—federal doors, especially.
The Pitch: Stop Scrubbing Through Hours of Video
Conntour's premise is straightforward enough. Security operators shouldn't need to manually scan hours of surveillance footage or navigate labyrinthine analytics dashboards to find what they're looking for. Instead, they should be able to type something like "man with sneakers passing a bag in the lobby" and get an answer.
The technical architecture underneath is more complex. The platform routes natural-language queries through multiple AI models and logic layers, processing video feeds either on-premises, in the cloud, or in hybrid configurations. The company claims it can handle up to 50 camera feeds on a single Nvidia RTX 4090 consumer GPU—a spec designed to appeal to customers with strict data sovereignty requirements who can't or won't send sensitive video to external servers.
That on-premises capability isn't incidental. It's central to Conntour's positioning in markets where "the cloud" remains a non-starter: intelligence agencies, law enforcement, defense contractors, critical infrastructure operators.
Traction in Unlikely Places
The customer roster remains mostly confidential, which is typical for this sector. One confirmed name: Singapore's Central Narcotics Bureau. Several others are government entities or publicly traded enterprises, according to Goldner, though he declined to specify further during a recent interview with TechCrunch.
The company also landed a spot in the inaugural cycle of Palantir's Startup Fellowship—a signal, if not quite an endorsement, of alignment with the defense-tech establishment. Bob Flores, former chief technology officer at the CIA, is an advisor. Conntour has been making the rounds at industry gatherings like Homeland Security Week, building relationships that could eventually translate into federal contracts.
Growing Quietly, With Global Footprints
Conntour emerged from Y Combinator's Winter 2025 batch and now employs between 11 and 50 people, with 14 based in Tel Aviv according to local reports, along with additional offices in Dubai and Singapore. The geographic spread reflects both the founders' backgrounds—Goldner and Kulla are based in Miami but maintain strong ties to Israel's tech ecosystem—and the global nature of their target markets.
The fresh capital will fund platform expansion and sales pushes into intelligence, law enforcement, defense, and transportation sectors. These are markets where AI-powered video analytics have become expected rather than novel, but where flexible query capabilities and on-premises control still function as meaningful differentiators.
Whether Conntour can convert early traction into sustained growth will depend partly on execution and partly on broader tailwinds. The convergence of generative AI, edge computing, and heightened concerns about critical infrastructure security has created an opening. How wide that opening proves to be—and how quickly competitors arrive—remains to be seen.
For now, though, the 72-hour fundraise tells its own story. Investors are paying attention.
