AI-powered civic assistant

🌍 About the Project: CivicPulse – AI-Powered Civic Assistant

💡 Inspiration

Every day, millions of citizens see issues like potholes, overflowing garbage, or broken streetlights — but most go unreported.
We were inspired by the idea of giving every citizen, regardless of literacy or language, a voice in governance.
CivicPulse empowers people to report civic problems using photo, voice, or text, while ensuring transparency and accountability through data and gamification.

⚙️ How We Built It

CivicPulse is designed as a modular, inclusive, and scalable system:

• Citizens report issues via photo, text, WhatsApp, or voice (IVR).
• AI Engine (image + voice + text models) classifies issues in real-time.
• Predictive Heatmaps highlight problem hotspots for early action.
• Admin Dashboard helps officials visualize and prioritize reports.
• Gamification Layer gives users “Civic Hero” scores for verified reports.

We followed a microservices architecture to build, integrate, and test features quickly — ideal for hackathon MVPs.

🚀 What We Learned

• Designing for inclusivity is just as important as designing for technology.
  
• Integrating AI with voice & language models taught us the importance of lightweight, mobile-friendly inference.
  
• Strong teamwork, modular coding, and efficient communication were key to finishing the MVP fast.

⚔️ Challenges We Faced

• Handling low-quality or noisy images/audio in real environments.
  
• Achieving multi-language accuracy for voice inputs (Hindi, English, and regional dialects).
  
• Maintaining data privacy and transparency simultaneously.
  
• Balancing AI accuracy and real-time responsiveness in limited infrastructure.

🌱 What Makes It Unique

CivicPulse bridges the citizen–government gap using AI + accessibility.
It not only enables people to report civic issues but also predicts and prevents future ones, driving data-backed, transparent, and inclusive civic action.

Built With

• celery
• distilbert-(text-classification)-??-**cloud-/-devops:**-aws-lambda
• distilled-whisper-(speech-to-text)
• fcm
• firebase/supabase-?-**ai/ml-models:**-yolov8n-(image-detection)
• frontend:**-vue-3
• ivr
• kubernetes-(k8s)
• litedb-for-mobile-caching-?-**apis-&-messaging:**-whatsapp-cloud-api
• mapbox
• notifications
• postgis
• push
• quasar-framework-(pwa)
• recharts/echarts-??-**backend:**-fastapi
• redis