Coram AI has secured $35m aimed at transforming wall-mounted security cameras into independent investigative tools driven by artificial intelligence.

The Series B round was co-led by new backer Ansa Capital and Battery Ventures, with participation from UP Partners, 8VC and Mosaic Ventures. The round brings total funding for the San Francisco firm to $66m.

Coram argues that physical security practices remain outdated. When incidents occur, teams often spend hours reviewing video, logs and alerts to reconstruct events.

The company offers software called Deep Investigation, an AI agent that accepts natural-language queries. It reviews months of footage, access records and visitor information from hundreds of cameras and locations, then delivers a compiled report. Tasks previously requiring hours now finish in minutes, according to the firm.

Established four years ago by Ashesh Jain and Peter Ondruska, Coram currently operates across more than 1,500 sites including schools and manufacturing facilities.

Privacy Considerations Amid Surveillance Capabilities

Coram emphasizes privacy protections. Its hardware runs AI models locally on NVIDIA processors at the edge, keeping sensitive video on premises rather than sending it to the cloud. The system also integrates with existing IP cameras, eliminating expensive hardware replacements.

At the same time the platform provides facial recognition, license-plate reading, tailgating alerts and real-time gun detection, with deployments at schools, places of worship and offices.

A Dallas-area megachurch uses the system to monitor 30,000 attendees across eight locations. One high school adopted live weapon detection to replace older cameras. The gains in speed are tangible, as is the scope of coverage.

The balance between added safety and expanded monitoring is familiar in AI security. Autonomous agents intensify the dynamic because a system that investigates across every camera and entry point also observes continuously and forms its own assessments.

Competition for the Security Operating System

Coram belongs to a group of startups seeking to establish an operating system for a specific sector by layering AI agents onto existing workflows. The company anticipates that buildings will eventually host hundreds of background agents.

Investors see a genuine opportunity. Physical security remains one of the largest sectors still awaiting modern AI adoption, noted Allan Jean-Baptiste of Ansa Capital, while legacy providers continue to focus on hardware and dashboards rather than autonomous functions despite heavy spending on AI in other fields.

Current performance claims such as 10x effectiveness gains and hundreds of agents per facility remain forward-looking projections. With $66m raised and 1,500 sites already active, the company has resources to evaluate whether future buildings will indeed monitor themselves.