Skip to content

birongliu/Spendly

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Inspiration

Managing finances effectively can be a challenge, especially with unpredictable spending habits and lack of real-time insights. We wanted to create an AI-powered tool that simplifies budget tracking, provides smart recommendations, and enhances financial awareness. Spendly was born to help users take control of their spending in a seamless and interactive way.

Screenshots

Watch the video:

Watch the video

What it does

Spendly is an AI-powered smart spending tracker that allows users to:

  • Set a Budget: Users create an account and set their monthly budget. (e.g., $1000/month)
  • Log Expenses:
    • Receipt Scanning: Scan receipts using OpenAI Vision, and an LLM extracts and categorizes expense details.
    • Voice Entry: (Future Feature) Use Whisper Speech-to-Text to record expenses verbally. Users can edit and submit transactions.
  • Dashboard Insights:
    • A circular chart displays budget usage.
    • Expense categories (e.g., Food: $123, Transport: $50) are highlighted.
    • AI-powered Q&A chatbot provides personalized budgeting insights, trend analysis, and financial tips.

How we built it

  • Account & Budget Setup: Users set their monthly budget.
  • Expense Logging:
    • Computer Vision (OpenAI Vision API) for receipt scanning and automatic categorization.
    • Whisper API (planned) for speech-based expense entry.
  • Dashboard & Insights:
    • Built using Chart.js and react-chartjs-2 for interactive spending visualizations.
    • AI chatbot analyzes trends and provides personalized financial advice.
  • Tech Stack:
    • Frontend: React, Typescript
    • Backend: Python, Flask
    • APIs: OpenAI Vision API, Whisper API.

Tech Stack

Challenges we ran into

  • Accurate expense categorization: Fine-tuning the LLM to correctly categorize expenses based on receipt text was complex.
  • Real-time data visualization: Implementing a responsive dashboard with Chart.js required careful data structuring.
  • Speech-to-text processing: Whisper’s integration faced initial issues in handling different accents and noisy backgrounds.
  • Optimizing chatbot responses: Ensuring AI suggestions were truly helpful and personalized for each user.

Accomplishments that we're proud of

  • Successfully built an AI-powered expense tracker with real-time categorization.
  • Designed an intuitive dashboard that makes budget tracking effortless.
  • Integrated OpenAI Vision to automate receipt scanning with high accuracy.
  • Developed an AI chatbot that provides smart financial insights.

About

maintain finances effectiveness

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •