AI SafeSpot – Women Safety Risk Prediction System

Inspiration

Women safety remains a major concern in many regions, especially in developing countries. I was inspired to use data and artificial intelligence to help identify potentially unsafe areas and support better decision-making. By analyzing crime statistics, I wanted to build a system that can provide a simple and understandable safety risk score.

What it does

AI SafeSpot predicts the safety risk of a location based on crime statistics such as rape cases, kidnapping, assault, domestic violence, and more. The system takes these inputs and generates a risk score between 0 and 1, which is categorized into Low, Moderate, or High risk. It also provides visual insights such as maps and charts to help users better understand crime trends and high-risk regions.

How I built it

I built the project using a combination of machine learning and web technologies. First, I trained a model using a real-world crime dataset containing state-wise crime statistics. The model was trained to learn patterns between crime features and overall risk. After training, the model was saved and integrated into a FastAPI backend.

The backend exposes an API endpoint that receives crime data and returns a predicted risk score. The frontend was built using Streamlit, which allows users to input data, view predictions, and explore visualizations like maps and charts. The frontend communicates with the backend API to fetch real-time predictions.

Challenges I ran into

One of the main challenges was understanding how to connect the machine learning model with a backend API and then integrate it with the frontend. Handling data correctly, especially mapping dataset columns to input features, was also tricky. I also faced issues related to environment setup, dependency installation, and debugging API requests.

Accomplishments that I'm proud of

I successfully built a complete end-to-end system that integrates machine learning, backend APIs, and a user-friendly frontend. The application can generate real-time predictions and visualize crime data effectively. I am proud of creating a project that not only demonstrates technical skills but also addresses a real-world social issue.

What I learned

Through this project, I learned how machine learning models are trained and deployed in real applications. I gained hands-on experience with FastAPI for backend development and Streamlit for frontend development. I also understood how to structure a full-stack AI project and work with real-world datasets.

What's next for AI SafeSpot – Women Safety Risk Prediction System

In the future, I plan to improve the model by using more advanced algorithms and larger datasets. I also aim to integrate real-time data sources and GPS-based location tracking to provide more accurate and dynamic predictions. Additionally, I plan to enhance the UI/UX and possibly develop a mobile application to make the solution more accessible.

Built With

• fastapi
• folium
• jupyter
• machine-learning-(scikit-learn)
• matplotlib
• numpy
• pandas
• plotly
• python
• rest-api
• scikit-learn
• seaborn
• streamlit