WattWise

WattWise analyzes your energy bills with AI to show you exactly what you're paying for, how much you're using, and how to save money. Upload your bill, get instant insights - it's that simple.

Comment

WattWise: AI-Powered Energy Bill Analyzer

Inspiration

Energy bills are confusing — filled with cryptic charges, hidden fees, and unclear usage graphs. Most people simply pay them without knowing:

  • Why was my bill so high this month?
  • What exactly am I paying for?
  • How can I reduce my energy consumption?

WattWise was built to answer these questions.
Our vision: Make energy bills as easy to understand as a bank statement.

What It Does

WattWise instantly analyzes any energy bill (PDF, scanned copy, or photo) and provides:

  • Clear cost breakdown of charges and fees
  • Usage insights across daily, monthly, and seasonal patterns
  • Carbon impact in real-world equivalents (miles driven, trees planted)
  • Personalized recommendations for energy savings

How We Built It

Tech Stack

  • Frontend: Next.js 13, TypeScript, Tailwind, shadcn/ui
  • Backend AI: Python Flask API with Ollama (Llama 3.2)
  • PDF & OCR: pdfplumber, EasyOCR
  • Deployment: Render (frontend + backend)

Architecture Highlights

  • Multi-layered PDF parsing (text-first, OCR fallback)
  • AI-powered recognition across utility companies
  • Lightweight CSS-based charts (optimized for memory limits)
  • Graceful fallbacks when AI or services are unavailable

Challenges We Overcame

  1. Messy PDFs – Some had selectable text, others were scanned images.
    Solution: Built a layered extraction pipeline: pdfplumber → OCR → server fallback.

  2. Memory Limits – Render’s free tier couldn’t handle heavy libraries.
    Solution: Replaced Chart.js with CSS visualizations, added code-splitting.

  3. AI Reliability – Ollama could be slow or unavailable.
    Solution: Added caching, mock responses, and graceful degradation.

  4. Bill Format Chaos – Every utility company had unique formats.
    Solution: Hybrid parsing system with regex + ML adaptation.

Key Features

  • Cost breakdown with simple visuals
  • Usage patterns to identify peaks and seasonal trends
  • Carbon footprint with real-world equivalents
  • Savings recommendations tailored to user habits
  • Accessibility: mobile-friendly, WCAG compliant, sub-3 second load

What We Learned

  • AI integration requires resilience and fallbacks
  • Performance optimization is critical under resource constraints
  • Sustainability data must be clear, relatable, and actionable

Future Vision

  • Smart home integration with IoT devices for real-time monitoring
  • Community features for comparing usage with neighbors
  • Predictive analytics to forecast bills and suggest optimal usage times
  • Direct partnerships with utility providers
  • One-click carbon offsetting through verified programs

Impact

  • Cost Savings: 15–25% reduction in average household bills
  • Environmental Awareness: Clear understanding of carbon impact
  • Peace of Mind: No more surprise charges or confusing bills

Built with passion at Hackathon for a more sustainable future.

Links

Built With

Share this project:

Updates