Meridian AI
European-engineered AI for intelligence and security teams worldwide. Analyse cases and documents, transcribe radio traffic and detect people, vehicles and objects in video. Existing capabilities, configured for your own infrastructure. Your data and decisions remain under your control.
- Role
- Second product
- Primary user
- Analysts, control rooms & security teams
- Core value
- Case and document analysis, source-backed answers, radio transcription and video analytics on your infrastructure. People review findings and decide.
Existing AI capabilities. A clear next step.
AI-assisted intelligence analysis, radio transcription and AI video analytics are existing Meridian capabilities. Specialised AI analysts that continuously correlate connected sources are our next product direction. Models, data access and integration are agreed for each deployment.
AI-assisted intelligence analysis
Examine authorised case files and documents, ask questions with source references, explore links between events and prepare intelligence summaries using local models. People verify the findings.
Explore this areaAI video analytics
Detect people, vehicles and objects in video and flag predefined anomalies for operator assessment. Models, video sources, hardware and integration are configured for the application.
Explore this areaSpecialised AI analysts
Continuously monitor connected information, investigate relationships and propose evidence-backed next steps. This is our development direction, not functionality offered as production-ready.
Explore this areaAn answer is a starting point. Understanding needs context.
Relevant information is spread across case files, documents, radio messages and footage. Meridian AI helps analysts examine that material, answer questions with sources, suggest links between events and prepare intelligence summaries. Video analytics adds detections for operator assessment. Your team checks the findings and decides how to act.
AI on your infrastructure
Run open-weight language models inside your own environment. Model selection, hardware and approved data access are configured to your operational requirements.
Speech to text for radio traffic
Push-to-talk messages are transcribed locally, so radio traffic becomes searchable and part of the case record.
Case and document analysis
Ask questions about authorised case files and documents in plain language. Examine relevant information and receive answers with source references your team can verify.
Suggestions, human decision
Meridian AI suggests links between events and drafts intelligence summaries for briefings. Your team checks the sources, reviews findings and decides on any follow-up action.
From video to operational information
Camera and drone footage becomes more useful when linked to a location, infrastructure and incident. Meridian can bring that footage together with the information your teams work with.
We configure video analytics to detect people, vehicles and objects, and flag predefined anomalies for operator review. Models, hardware, approved video sources and links to incident context are validated for each application.
AI video analytics is an existing Meridian capability. Model performance, hardware and integration are validated against the agreed operating requirements. Operators review detections and decide on follow-up. Integration into drone operations remains on our roadmap.
Compare illustrative video examples
Illustrative scenario: detection boxes mark a moving van, a forklift and containers. The van links to fictional vehicle record DEMO-001, pending operator review. AI video analytics is an existing capability; this clip uses generated footage and illustrative tracking, not a live system recording.
Beyond data. Towards continuous intelligence.
Our vision is to empower organizations with specialized AI analysts that continuously monitor connected information, uncover relationships, identify emerging developments and recommend meaningful next steps.
Designed to augment human expertise. Human analysts remain responsible for important decisions and sensitive actions.
01Connected data
RSS / news feeds
Network disruption reportedERP / enterprise systems
Maintenance window loggedSovereign / private cloud
Service region contextOperational databases
Infrastructure recordsExternal APIs
Service alert received02—04Correlation & analysis
Meridian AI
Ontology · skills · permissions · policies
05Actionable intelligence
Potential incident detected
A reported network disruption may be related to scheduled maintenance in the same service area.
Possible correlation · not yet verified
Review example evidence
- [01]News feed: a service disruption is reported in the northern service area.
- [02]Enterprise record: planned maintenance affects the same service area.
- [05]External API: a service alert overlaps the reported maintenance window.
Suggested next steps · analyst approval
- Open an investigation
- Prepare an incident report
- Notify the authorized team
06For human review
Review the evidence. Decide the next step. No action is taken automatically.- 01Connected data
- 02Continuous ingestion
- 03Meridian AI
- 04Correlation & analysis
- 05Actionable intelligence
- 06Human analyst
Specialized by domain
- Police intelligence
- Defence intelligence
- Telecom intelligence
- Critical infrastructure
Your environment. Your AI. Your decisions.
Meridian AI is designed to keep model inference, case context and results within your controlled infrastructure. Model selection, source access, permissions and integration boundaries are defined for each deployment. Specialised analysts must build on that same foundation of control and human responsibility.
Data architecture: your own data layerEvery part of Meridian is designed on the same sovereign architecture: European engineering, customer-controlled keys and no telemetry — for your own cloud, on-premise or air-gapped environment. Deployment is scoped to your operational and applicable legal requirements.
Sovereignty
