Tweaker-Tea

• 

Inspiration

Walking home late at night can feel unpredictable, especially in busy downtown areas. Many people rely on instinct when deciding which streets or alleys to avoid, but there’s often no easy way to see where recent incidents have occurred.

Cities already publish crime statistics and safety reports, but this information isn’t typically integrated into tools people actually use when navigating. We were inspired by navigation apps that use real-time data to improve decision-making, and we asked a simple question:

What if people could see a safety heatmap of the city before choosing how to walk home?

That idea led to Tweaker-Tea.

What it does

Tweaker-Tea is a safety-focused navigation tool that helps pedestrians avoid potentially dangerous areas when traveling through the city.

The app overlays a heatmap of recent crime activity across cities like Toronto and Kitchener-Waterloo. Instead of simply choosing the fastest route, users can see which streets have higher reported incidents and choose safer paths.

Key features include:

• A crime heatmap generated from publicly available police reports • Crowdsourced safety reports from users in real time • A safe route recommendation system that avoids high-risk areas • A panic button that can alert friends and family with live location • A loud alarm and flash feature to draw attention during emergencies

Our goal is simple: help people make more informed decisions when walking through unfamiliar or unsafe areas.

How we built it

We built Tweaker-Tea by combining open crime data with mapping and visualization tools.

First, we collected publicly available crime reports from the Waterloo Regional Police open data portal. These reports were processed and mapped into geographic coordinates to generate a city-wide heatmap of reported incidents.

Next, we layered this data on top of a map interface and implemented a routing system that identifies paths avoiding high-risk areas.

The system calculates safety scores for different streets based on incident density and visualizes these scores using a color-coded heatmap.

We also added emergency safety tools, including a panic button and location-sharing functionality, to help users quickly alert trusted contacts if they feel unsafe.

Challenges we ran into

One of the biggest challenges was working with real-world crime data. Public safety data can be inconsistent in format, incomplete, or difficult to translate into usable geographic information.

Another challenge was balancing route safety with route efficiency. If the algorithm avoided every area with any crime report, routes could become unrealistic. We had to experiment with weighting systems to find a reasonable balance between safety and practicality.

Finally, we had to design a system that was simple enough for users to understand quickly while still presenting meaningful safety insights.

Accomplishments that we're proud of

Within the hackathon timeframe, we successfully built a working prototype that:

• Visualizes crime data as a dynamic city heatmap • Demonstrates safer route planning based on incident density • Integrates emergency safety tools such as location alerts and alarms

We're proud that we were able to turn publicly available data into something that could realistically help people make safer decisions.

What we learned

This project taught us a lot about working with open datasets and converting raw information into meaningful insights.

We also learned how important data interpretation is when dealing with safety-related information. Simply displaying data is not enough; it needs to be presented in a way that is understandable and responsible.

Finally, we learned how much impact even a simple prototype can have when it addresses a real-world concern.

What's next for Tweaker-Tea

If we continued developing Tweaker-Tea, we would focus on expanding both the data sources and the intelligence of the system.

Future improvements could include:

• Real-time user safety reports and verification systems • Machine learning models that predict risk based on time of day and location • Integration with more police open-data portals across Canada • Improved safe-route algorithms that account for lighting, foot traffic, and transit access • Partnerships with universities to help students navigate safely at night

Our long-term goal would be to build a Waze-style safety navigation platform for pedestrians, helping people move through cities with greater awareness and confidence.

Businesses would pay us in advertisement revenue to be listed as "safe spots" as the user progresses through the map similar to Waze!

MLH x ElevenLabs — Best Project Built with ElevenLabs — You're using their Conversational AI agent ("Toronto Mans" voice chat). This is your strongest track fit.

Google — Build with AI Track — If you integrate Gemini for route safety analysis or incident prediction, this fits well.

[MLH] Best Use of Gemini API — Same as above, overlaps with Google track.

[MLH] Best Use of Auth0 — You have @auth0/auth0-react installed. If auth is wired up, this qualifies.

SPUR Founder Track — Build a Real Canadian Startup — Safety navigation for pedestrians is a real Canadian problem with startup potential.

Vivirion Solutions — Best Practical Healthcare Hack — Stretch, but pedestrian safety/emergency tools could loosely connect.

Most Technically Complex AI Hack — If you layer in AI-powered route prediction or agentic safety analysis.

Built With

• 18
• auth0
• cmdk
• css
• dom
• elevenlabs
• form
• framer
• hook
• leaflet.heat
• leaflet.js
• motion
• query
• radix
• react
• recharts
• router
• shadcn/ui
• sonner
• tailwind
• tailwindcss-animate
• tanstack
• typescript
• ui
• vaul
• vite
• zod