RAKSHIKA AI

• 

• 

• 

• 

• 

• 

Inspiration

Women’s safety is still a major concern, especially when someone is travelling alone, returning home late, or facing an unexpected emergency. Existing safety solutions often depend on the user manually pressing an SOS button after a dangerous situation has already occurred.

We wanted to build something more proactive than reactive.

This inspired us to create RAKSHIKA AI — an AI-Powered Women’s Safety & Guardian Ecosystem that can continuously understand safety signals, detect potential danger, and connect the user with trusted guardians when help may be needed.

What it does

RAKSHIKA AI is a single application with two connected modes:

👩 Women’s Mode

Designed for the person carrying the device.

Key features:

One-tap SOS emergency alert

Live location sharing with trusted contacts

AI-powered danger-word/audio detection

Camera and motion-based emergency detection

Automatic emergency evidence capture

Trusted emergency contacts

Safety status and active protection monitoring

Quick emergency calling and messaging

🛡️ Guardian Mode

Designed for trusted family members or guardians.

Guardians can:

View shared live location

Receive emergency alerts

Monitor safety status

Receive AI-generated risk notifications

Access emergency information shared by the user

Respond quickly during a potential emergency

The core idea is to create a connected safety ecosystem instead of just an SOS button.

How we built it

RAKSHIKA AI was designed as a modular, mobile-first safety platform.

The system consists of:

User Layer → Women’s Mode and Guardian Mode

AI Safety Layer → Audio/danger-word detection, motion analysis and risk assessment

Emergency Layer → SOS, location sharing, emergency contacts and alerts

Guardian Layer → Real-time safety updates and emergency response

Privacy Layer → User-controlled permissions, consent-based location sharing and controlled access to sensitive information

We focused on creating a simple interface so that emergency actions can be performed quickly, even under stress.

Challenges we ran into

The biggest challenge was designing a system that could be powerful enough to detect potential danger while remaining simple and reliable during emergencies.

Major challenges included:

Detecting potential danger without generating too many false alarms

Designing AI-based audio and motion detection for real-world situations

Creating fast emergency communication and location sharing

Making the interface usable during stressful situations

Balancing continuous safety monitoring with user privacy

Ensuring that sensitive information is only shared with authorized guardians

Designing both Women’s Mode and Guardian Mode as one connected ecosystem

Making the concept practical for real-world deployment rather than just a demonstration

Accomplishments that we're proud of

We are proud that RAKSHIKA AI goes beyond traditional emergency applications.

Our key accomplishments include:

Built a dual-mode safety ecosystem with Women’s Mode and Guardian Mode

Designed proactive AI-based danger detection

Integrated live location sharing

Created an emergency SOS workflow

Added audio/danger-word detection concepts

Added camera and motion-based emergency detection

Created connected guardian notifications

Designed a privacy-focused permission model

Created a simple, emergency-friendly user interface

Connected multiple safety signals into a unified risk-assessment concept

Most importantly, we transformed the idea of women’s safety from “press SOS when something happens” into “detect risk early and connect help faster.”

What we learned

While building RAKSHIKA AI, we learned that technology for safety must prioritize reliability, simplicity and privacy over unnecessary complexity.

We learned:

AI detection must be designed around real-world false positives and false negatives.

Emergency interfaces should require minimum interaction.

Privacy and user consent are essential when handling location, audio and camera data.

A safety application becomes more useful when it connects the user with a trusted support network.

AI should assist humans rather than completely replace human judgment in critical situations.

Real-world safety products need strong testing across different environments and scenarios.

The biggest lesson was:

“In an emergency, technology should reduce the time between detecting danger and receiving help.”

What's next for RAKSHIKA AI

Our long-term vision is to transform RAKSHIKA AI into a proactive personal safety ecosystem.

Phase 1 — Smarter Detection

Improved AI audio detection

Better motion and activity recognition

Context-aware risk scoring

Reduction of false alarms

Phase 2 — Faster Emergency Response

Automated emergency escalation

Trusted guardian networks

Emergency response coordination

Location-based emergency assistance

Phase 3 — Privacy & Security

On-device AI processing wherever possible

Stronger consent controls

Encrypted emergency data

Temporary access to sensitive information

Privacy-preserving safety monitoring

Phase 4 — Intelligent Safety Assistant

An AI safety companion that can understand context such as:

“I’m travelling alone.”

“My route suddenly changed.”

“I haven't reached my expected destination.”

The system could then provide proactive safety recommendations or escalate to trusted contacts based on user-defined policies.

Phase 5 — Wider Ecosystem

RAKSHIKA AI could eventually integrate with:

Wearable devices

Smartwatches

Public safety infrastructure

Transportation platforms

Emergency services

Campus safety systems

Verified community responders Final Vision:- RAKSHIKA AI aims to become more than an emergency app — it aims to become an intelligent, privacy-conscious safety companion that helps detect risk early, connect trusted people quickly, and give women greater confidence to move freely.

Built With

• ai
• css3
• html5
• javascript
• qrickit-qr-code
• react
• tailwind
• typescript
• vite