Core Functionality

Alia is a personal safety application that provides women with an instant AI-powered voice companion during moments of vulnerability. With a single tap, users initiate a phone call with an adaptive AI agent that dynamically assesses the situation through natural conversation. The AI operates across multiple (escalating) response tiers: companionship mode (staying on the line while walking home), guidance mode (directing users to safe locations, nearby businesses, or well-lit routes), and emergency mode (connecting to local emergency services with real-time location data). The app disguises itself as a regular contact, allowing users to say “I’m calling Alia” without raising suspicion.

Working Prototype & Technical Complexity

Voice AI: Real-time speech recognition and natural language processing to understand context, detect stress levels in voice patterns, and respond conversationally
Location Services: GPS tracking with points-of-interest awareness (police stations, hospitals, 24-hour businesses, safe havens)
Tool Orchestration: The agent can autonomously trigger actions: sending location shares to emergency contacts.

Innovation & Creativity

Alia reimagines personal safety by replacing panic buttons with calm conversation. Rather than escalating fear, the AI provides graduated support, meeting users where they are emotionally. The “alias” concept itself is innovative: safety technology that hides in plain sight.

Real-World Impact

One in three women experiences gender-based violence globally. Alia provides an always-available companion that requires no explanation to bystanders or potential threats, just a woman talking to a friend.

Theme Alignment

Alia exemplifies the hackathon’s conversational agent theme by architecting a fully autonomous safety assistant built on multi-agent orchestration.

Conversational AI at the Core

The primary interface is a real-time voice agent (not a button press or text menu). Users interact through natural speech, and the AI adapts its conversational style based on detected urgency, emotional state, and context.

Multi-Agent Architecture

Alia coordinates specialized agents working in concert:

• Assessment Agent: Continuously analyzes voice patterns, keywords, and context to determine threat level
• Navigation Agent: Queries location APIs and points-of-interest databases to identify safe destinations
• Escalation Agent: Monitors conversation state and autonomously triggers emergency protocols when thresholds are met
• Communication Agent: Manages parallel tasks: alerting emergency contacts, preparing location data for dispatchers, maintaining the user-facing conversation

Partner Technology Integration

The solution leverages multiple partner technologies: voice AI infrastructure for low-latency speech processing, cloud-based LLM orchestration for contextual reasoning, telephony APIs for emergency service bridging, and geolocation tools for real-time routing, all unified through a single conversational experience.

Autonomous Decision-Making

Critically, Alia doesn’t wait for explicit commands. The agent system independently escalates responses, shares locations, and initiates emergency calls when the situation warrants, transforming from passive tool to proactive guardian.