Launches AI-Powered Forensics

New forensics capabilities and intuitive search across the customers’ entire network enable near-real-time investigations. launched its AI-Powered Forensics solution to power near real-time incident investigations. New entity and event-based search capabilities allow teams to search by objects, complex behaviors, and non-biometric person identifiers such as shirt color—reducing investigation times by over 90%.

Today, security investigations take hours or even days and require operators to review individual camera streams with limited search options (typically just motion or basic person and object search), log into numerous applications, and wait for time-consuming video processing. Ambient’s AI-Powered Forensics is fully integrated with the camera system enabling near-instant search across camera networks after an incident has occurred—eliminating prolonged wait times and processing so security investigations are reduced significantly to mere minutes, rather than hours or days.

Entity Search builds on’s event-based search and filtering capabilities, leveraging non-biometric identifiers for objects and people to pinpoint key footage along with complex event signatures (i.e.: a person in a black shirt removing a package). Additionally, Conjunction Search opens up new possibilities for search parameters by allowing users to filter footage by two or more events, entities, or attributes.

One-click sharing and archiving capabilities allow for the immediate delivery of footage across teams to support expedited internal and external investigations, build detailed reports on incident timelines and response, and ensure compliance with security protocols. Streamlined investigations will allow security teams to analyze footage to identify the incidence rate of specific threats, understand the efficacy of security measures, and take measures to address vulnerabilities.

The new AI-Powered Forensics leverages the Context Graph, which uses computer vision intelligence to identify entities, objects, and their interactions across surveillance feeds and PACS alarms. The extensible graph powers the Platform’s suite of signals intelligence, threat detection, and response capabilities. To continuously strengthen performance, regularly adds new event and entity recognition signatures while utilizing human-in-the-loop feedback for training and accuracy.

Click here for more product news in Building Automation.