Tue Nov 05 2024

PerCent – AI-Powered Personal Finance Assistant 💰
Introduction
PerCent is a personal finance assistant leveraging AI-driven insights and goal tracking to simplify financial planning. It combines a RAG-based AI chatbot with robust backend technologies, empowering users to make informed financial decisions while managing expenses, savings, and investments.
Motivation
Millions struggle with managing their finances, especially young professionals and students transitioning to financial independence. PerCent addresses this challenge by offering a simple, personalized tool to streamline budgeting, track goals, and gain financial literacy.
Problem Definition
The project solves critical financial challenges:
- Expense Tracking: Simplifies categorization and analysis of spending.
- Goal Tracking: Helps users define and monitor short- and long-term financial goals.
- Real-Time Financial Insights: Leverages RAG workflows to deliver accurate advice via AI chatbots.
Functional Requirements
- Goal Setting: Users can set monthly or yearly savings goals and track progress.
- Expense Categorization: Automatically categorizes income and expenses for better insights.
- AI Chatbot: Provides personalized financial advice through conversational AI.
- Financial Literacy Hub: Offers curated courses and financial news.
Non-Functional Requirements
- Performance: Supports real-time response for 10k+ concurrent users.
- Scalability: Designed to handle increasing user loads and additional features.
- Security: Ensures encrypted transactions and protects sensitive data.
Technology Stack
Technology | Purpose |
---|---|
React Native | Cross-platform app development |
React.js | Web-based financial dashboard |
Express.js | Backend API framework |
PostgreSQL | Relational database for user data |
Redis | Caching and message queue implementation |
RabbitMQ | Message broker for asynchronous tasks |
Hugging Face | NLP models for semantic search |
Qwen-14B-Chat | AI model for chatbot responses |
Docker | Containerized deployment |
AWS Lambda | Serverless computing for AI pipelines |
CloudWatch | Monitoring and logging system metrics |
Polygon | Blockchain integration for financial services |
Methodology
Workflow Architecture
- Frontend Interaction:
- Users interact with the app through mobile (React Native) or web dashboards.
- Message Queue:
- Tasks are queued via RabbitMQ or Redis Streams for asynchronous processing.
- AI Insights:
- Semantic search and summarization via Hugging Face models.
- NLP query generation and responses powered by Qwen-14B-Chat.
- Backend Processing:
- Expense and goal data are processed using Express.js and stored in PostgreSQL.
- Real-Time Notifications:
- Updates are sent to the frontend via Redis Pub/Sub and WebSockets.
Features
- AI Chatbot:
- Provides personalized advice, such as spending optimizations or investment recommendations.
- Goal Tracking:
- Users can set goals, monitor progress, and receive actionable insights.
- Expense Categorization:
- Automatically organizes expenses into categories with visual reports.
- Financial Literacy:
- Access courses, tutorials, and news to improve financial knowledge.
Assessment Framework
To ensure optimal performance and resource efficiency, PerCent includes a sustainability assessment framework:
- Message Queue Metrics:
- Evaluate task throughput and latency with RabbitMQ and Redis Streams.
- AI Model Performance:
- Analyze latency and accuracy of Hugging Face and Qwen models.
- Database Optimization:
- Monitor query execution times and caching efficiency in PostgreSQL and Redis.
Conclusion
PerCent delivers a seamless and AI-driven approach to financial planning, empowering users to manage their finances with ease. It effectively integrates real-time insights, goal tracking, and expense management while leveraging cutting-edge AI and scalable technologies.
The project aligns with:
- SDG 9: Industry, Innovation, and Infrastructure.
- SDG 12: Responsible Consumption and Production.
References
- Brown, L., et al. (2023). AI and Financial Empowerment: Semantic Search in Finance. Journal of Applied AI Research.
- Hugging Face (2024).
all-mpnet-base-v2
Semantic Search Documentation. - RabbitMQ (2023). RabbitMQ Guide to Message Queue Optimization.
Future Enhancements
- Predictive Analytics: Integrate AI-powered forecasting for savings and investments.
- Family Integration: Allow shared financial goals for collaborative planning.
Team
- Frontend & UX: Aditya Kulkarni
- Backend & AI Integration: Aditya Pai
GitHub Repository
Thank you for exploring PerCent! We are committed to helping individuals achieve financial independence through innovative technology and AI-driven insights. 😊