Your voice.
Your safety.
Your Guardian.
GuardianVoice NG helps detect distress signals, confirm emergency intent, and route alerts through the configured response workflow.
One tap to start
Trigger the alert flow and review the live status, location, and rescue coordination steps.
Live user position
How it works
Listen
Detect a distress phrase or unusual sound pattern from the user’s microphone.
Assess
Check the severity and confirm whether the incident meets emergency criteria.
Respond
Send alerts and share location with the configured contacts or responders.
From Distress Signal to Coordinated Response
GuardianVoice balances rapid detection with deliberate safety verification to avoid false alarms.
Voice Detection Stage 2 Active
GuardianVoice listens for emergency signals, distress phrases such as Help me or Emergency, and sudden acoustic loudness spikes.
AI Analysis AI Evaluator
The system analyzes voice input and acoustic patterns for possible emergency indicators and computes a threat confidence score.
Confirmation Window Safety Gate
A rapid safety countdown lets the user dismiss an accidental trigger before an alert is dispatched.
Location Anchoring Device Dependent
With permission, browser-provided coordinates and timestamp metadata can be attached to an emergency packet.
Emergency Alert Dispatch Contact Opt-In
An alert can be routed to configured emergency contacts and authorized responders.
Coordinated Response Dashboard
The response dashboard displays active incidents, coordinates, status updates, and alert history.
Built for rapid response
AI Voice Detection
Browser microphone monitoring with Web Audio API and Speech Recognition to capture distress phrases.
Distress Detection
Acoustic analysis measures sudden shouting and panic signals to distinguish potential distress from background noise.
Emergency Confirmation
A safety grace window helps prevent false alarms and lets users cancel accidental triggers.
GPS Location
Device positioning uses the accuracy reported by the browser and requires location permission.
Real-Time Alerts
Emergency notifications can be sent to designated contacts through the configured SMS service.
Emergency Response Dashboard
A centralized view for active emergency incidents, location telemetry, and response status.
Alert History
Filter and search past emergency activations, timestamps, detection methods, locations, and outcomes.
Mobile Support
Responsive layouts support smartphones and tablets with touch-friendly controls.
Safety Monitoring
Telemetry displays system readiness, microphone status, location availability, and network state.
Built to help fast and stay aware.
GuardianVoice NG is a safety-first platform for voice distress detection, alert coordination, and emergency monitoring workflows.
Empowering Every Voice
To explore accessible technology that can help people communicate possible emergency situations faster.
When sudden danger limits access to a phone screen, speech-activated safety tools can help people request assistance through everyday devices.
Engineering Practical Safety
To demonstrate how web and AI technologies can be combined into practical safety-focused tools.
The project brings together client-side audio analysis, progressive web technology, and clear alert coordination.
Core Technology Stack & Pillars
Built on accessible web engineering standards.
Artificial Intelligence
Pattern matching and semantic distress heuristics.
Voice Recognition
Web Audio API and browser speech recognition.
Location Technology
Browser geolocation with accuracy reporting.
Web Applications
HTML5, CSS3, and Vanilla JavaScript.
Real-Time Databases
Event queues and response state tracking.
Send an inquiry
Share feedback, request a demonstration, or ask about the research initiative.
Contact GuardianVoice NG
Emergency Notice
GuardianVoice NG is an academic prototype and does not replace official emergency dispatch. In a life-threatening situation, contact emergency services directly.
Student Research Initiative
A collaborative project exploring practical applications of AI, speech recognition, and responsive UI design for public safety.
Open Web Standards
Built with HTML5, CSS3, and modern Vanilla JavaScript for accessibility and educational clarity.