ALICE: AI-Powered Intelligent Personal Finance Assistant Platform

ALICE is a modern fintech platform integrating generative AI and predictive machine learning, designed to provide personalized financial management solutions for young users. Its core functions include an AI chatbot, budget optimization, balance prediction, impulse spending alerts, and user grouping analysis. The project aims to help young people better manage their finances through intelligent, real-time guidance.

In today's fast-paced life, personal financial management is a challenge for many young people. ALICE (Artificial Intelligence for Literacy, Ideal Allocation, and Cost) was created to address this issue. It combines modern web technology, generative AI (large language models), and predictive machine learning to offer real-time, highly personalized financial guidance—going beyond a simple bookkeeping tool to act as an intelligent assistant that understands users' financial needs.

ALICE offers several key features:

1. AI Financial Assistant: Integrated with Google Gemini and Llama 3, it understands user financial profiles, analyzes transaction history, and provides tailored advice via natural language.
2. Budget Optimization: Uses DNN to analyze income/expenses, adjust budget strategies dynamically, and maximize savings.
3. Balance Prediction: LSTM model predicts 10-day balance trends to warn of potential fund shortages.
4. Impulse Spending Intervention: Identifies high-risk transactions and sends alerts to help build healthy habits.
5. User Grouping: Autoencoder classifies users into thrifty, moderate, or high-impulse groups for precise recommendations.

ALICE uses a microservices architecture:

• Frontend: React 19 + Vite (TypeScript, TailwindCSS v4, TanStack React Query/Axios) for smooth UI.
• Backend Gateway: Node.js + Express (PostgreSQL) for user auth, data storage, and context provision.
• AI Chatbot Service: Python + FastAPI (Google Gemini as main model, Groq/Llama-3 as backup) with dynamic prompt engineering.
• Predictive ML Service: Python + FastAPI + Keras/TensorFlow (LSTM, DNN, Autoencoder) for ML inference.

Differentiated deployment strategies are suggested:

Microservice	Recommended Platform	Notes
Frontend	Vercel	Perfect integration with Vite
Backend	Vercel/Render	Node.js serverless deployment
Chatbot	Vercel	vercel.json config provided
Predictive ML	Render.com	TensorFlow exceeds Vercel's 250MB limit, needs full server environment
This approach balances cost-effectiveness and resource needs.

ALICE's value lies in:

1. Lowering Financial Barriers: AI dialogue makes complex financial concepts accessible.
2. Prevention Over Remediation: Predictions and alerts help avoid financial crises.
3. Personalization: Tailored advice instead of one-size-fits-all solutions.
4. Behavior Change: Gentle nudges to foster healthy financial habits.

ALICE represents a fusion of fintech and AI. It acts as an intelligent assistant (not a decision-maker) providing information and advice. For developers, it demonstrates combining LLMs with traditional ML and microservices. For users, it predicts the future of personal finance tools: smarter, more proactive, and personalized.