Flaggr

• 

  current/escalating situation measure - panic button

• 

  preventative measure - fake call

Inspiration

Flaggr was inspired by traditional physical panic buttons and modern safety platforms like LiveSafe. While these tools provide reactive emergency support, we saw an opportunity to create a smarter, more preventative solution—one that not only responds to danger but actively monitors and documents escalating situations in real time.

What It Does

Flaggr is a preventative safety app designed to empower users in vulnerable situations.

At its core is a discreet panic button that immediately begins recording audio when activated. This creates reliable documentation in scenarios that often become “my word against yours” situations, giving users tangible leverage and protection.

Simultaneously, Flaggr analyzes the conversation in real time. If the system detects escalating language or threatening behavior, it can automatically alert authorities—removing the burden from the user during high-stress moments when manual intervention may not be possible.

How We Built It

Tech Stack:

• Speech-to-Text and Text-to-Speech: ElevenLabs
  
• AI Analysis: Google Gemini
  
• Frontend: React
  
• Development Environment: Vite
  
• Database: SQLite
  

This combination allowed us to rapidly prototype an intelligent, responsive safety system with full voice-processing capabilities.

Challenges We Faced

One of our biggest goals was to implement a long-term preventative feature: analyzing conversations directly within messaging platforms to detect unsafe situations before they escalate offline.

However, we ran into significant API restrictions—particularly with accessing conversation data from platforms owned by Meta Platforms. Messaging platform limitations made this integration infeasible within the hackathon timeline.

Additionally, due to time constraints and limited mobile development experience, we were unable to deploy Flaggr as a native mobile application—despite that being our ideal format for accessibility and real-world use.

Accomplishments We’re Proud Of

• Built a fully functional end-to-end application
  
• Implemented seamless speech-to-text and text-to-speech capabilities
  
• Integrated real-time AI-driven threat analysis
  
• Delivered a working prototype demonstrating both reactive and preventative safety features
  

What We Learned

We learned the importance of early planning—especially when working with third-party APIs outside of the hackathon’s core tools. Understanding access restrictions and technical limitations earlier would have helped us pivot faster and allocate development time more efficiently.

What’s Next for Flaggr

• Explore compliant and secure methods for integrating social platform APIs
  
• Develop a fully functional mobile application for easier real-world accessibility
  
• Refine AI threat-detection models for greater accuracy and reliability
  
• Expand preventative features beyond emergency response
  

Flaggr is just the beginning. Our long-term vision is to redefine personal safety through intelligent, proactive technology—empowering users before situations escalate, not just after.

Built With

• elevenlabs
• gemini
• python
• react
• sql