Inspiration In many cities, citizens raise complaints about issues like water leakage, garbage overflow, potholes, or streetlight failures — but the response process is often slow, manual, and lacks transparency. We noticed that complaints are usually collected, sorted, and assigned manually, which delays action. This inspired us to build a system that could automatically understand complaints, classify them using AI, and route them to the right department instantly.
Our vision was simple: Transform unstructured citizen complaints into real-time civic intelligence.
⚙ What it does AI Citizen Signal Engine is an AI-powered civic complaint management system that: Automatically classifies complaints using Machine Learning Detects and maps complaint locations Assigns complaints to the correct department Provides real-time dashboards for analytics Tracks resolution time and department efficiency Allows departments to update complaint status
It creates a seamless workflow from citizen submission → department processing → resolution tracking.
How we built it We built the system using: Streamlit for the interactive web interface Scikit-learn (TF-IDF + Logistic Regression) for complaint classification TextBlob for sentiment analysis and priority detection Pandas for data handling Plotly & Folium for interactive analytics and geo-mapping The AI model converts complaint text into numerical features using TF-IDF:
We implemented multi-role access: Citizen Panel Department Panel Admin Panel Each role has filtered real-time data visibility.
Challenges we ran into
Ensuring strict role-based data access (avoiding data leakage between departments) Managing real-time session state in Streamlit Training an accurate ML model with limited data Handling live location detection issues across devices Maintaining dashboard updates without performance lag These challenges pushed us to refine both backend logic and frontend flow.
Accomplishments that we're proud of Successfully integrated an ML model into a live civic dashboard Built a complete multi-role governance system Implemented real-time complaint analytics Designed a scalable architecture for smart city applications Created an end-to-end automated complaint routing system Most importantly, we transformed a social problem into a working technical solution.
What we learned
How to deploy machine learning inside web applications The importance of clean data filtering and access control Real-world debugging of session state and dynamic dashboards How AI can directly improve governance and public service delivery This project strengthened both our technical skills and our understanding of civic technology.
What's next for AI Citizen Signal Engine
Multilingual complaint support SMS & WhatsApp notification integration Advanced AI-based severity prediction Integration with Smart City and Government API Predictive analytics for urban planning Our goal is to evolve AI Citizen Signal Engine into a full-scale smart governance platform that empowers both citizens and administrations.
Built With
- datetime
- logistic
- pandas
- plotly
- python
- regression
- scikit-learn
- streamlit
- textblob
- tf-idf
- uuid
- vectorizer
Log in or sign up for Devpost to join the conversation.