SafeStride

User Reports and History Crime Data is plotted on a heatmap with increasing intensity denoted by color and route passes through fewer spots
Sign Up page - Takes User Details and Close point of contact for the user for emergency notifications.
Login Page for users to Sign in to SafeStride
User inputs a To and From destination and the most safest and optimal route is provided.
In the Report Section, users can report emergencies as and when they happen. Users record their audio stating the emergency they are in.
Once the user submits the recording, it is sent for processing into Gemini which identifies the type of Crime and stores it as a report.

Inspiration:

SafeStride was sparked by a widespread yet troubling reality: many women often feel unsafe when walking alone, particularly in unfamiliar areas at night. While current navigation apps optimize for speed or distance, they rarely prioritize safety—the very factor that matters most in these situations. Our goal was to change that narrative. SafeStride was created to bridge this critical gap, empowering users—especially women—with smarter, safety-centric navigation designed for peace of mind.

What It Does:

SafeStride is a safety-first navigation app that helps users identify the safest walking routes using real-time contextual data. It evaluates multiple route options by calculating a Weighted Safety Score based on: Historical crime data Real-time foot traffic Street lighting coverage Proximity to institutions like schools, hospitals, police/fire stations Community-reported incidents

To further enhance safety, SafeStride includes an AI-powered fake call feature using the Gemini API, simulating live conversations to deter suspicious individuals. It also offers a built-in community incident reporting tool, allowing users to stay informed and support each other.

How we built it

Google Maps API was used for live location tracking, route suggestions, and navigation. We developed a custom algorithm to compute safety scores by processing crime statistics, traffic patterns, and local infrastructure data from our in-house databases. The Gemini API powers the analyzes, summarizes and identifies the threat severity and sends the keywords to be included to the DB. Flask, Firebase, and MongoDB Atlas were leveraged for backend infrastructure and deployment.

Challenges we ran into

Striking the right balance in keyword-based emergency triggers—being both sensitive and accurate.
Designing a real-time safety score without overloading the user interface.
Developing AI-generated calls that feel realistic without using actual voices.
Ensuring secure, fast sharing of emergency locations while maintaining user privacy.

Accomplishments that we're proud of

Seamless integration of Google Maps with real-time safety overlays and route suggestions.
Successfully embedding the AI call system using Gemini API.
Delivering a thoughtful, intuitive user experience that feels discreet yet powerful.
Crafting a project that champions Tech for Good and supports Women’s Empowerment.

What we learned

How to extend traditional navigation APIs into safety-first applications.
The critical role of user psychology when building features meant for emergencies.
The technical finesse required to combine live data, emergency protocols, and intuitive UX into a single system.
The value of rapid prototyping and team collaboration in a high-pressure environment like a hackathon.

What's next for SafeStride

Introducing customizable fake call voices and scripted dialogues.
Adding a discreet SOS vibration or beeper system for quick signaling.
Expanding to smartwatch and wearable device support.
Launching a Guardian Mode, allowing trusted contacts to track your journey live.
Preparing for app store deployment and forging partnerships with local safety organizations and municipal data providers.