Track. Understand. Predict. Improve.
Fintra-AI is an intelligent personal finance management platform designed to help users track expenses, understand spending behavior, manage budgets, and make data-driven financial decisions.
By combining modern full-stack web technologies, Artificial Intelligence, Machine Learning, and secure cloud infrastructure, Fintra-AI transforms raw financial data into meaningful insights and actionable recommendations.
- Overview
- Why Fintra-AI?
- Key Features
- AI & Machine Learning
- Analytics
- Security
- Technology Stack
- System Architecture
- Project Structure
- Getting Started
- Environment Variables
- Database Setup
- Development Workflow
- Documentation
- Roadmap
- IEEE Summer of Code 2026
- Contributing
- Future Vision
- License
- Maintainer
Fintra-AI is an AI-powered personal finance management platform designed to transform raw financial data into actionable insights.
Instead of simply recording transactions, Fintra-AI aims to help users understand where their money goes, why they spend it, what may happen next, and how they can improve their financial health.
The platform combines:
- 💰 Personal finance management
- 🤖 Generative AI
- 🧠 Machine Learning
- 📊 Financial analytics
- 🔐 Secure authentication
- ☁️ Cloud-native infrastructure
- 📈 Predictive financial insights
Explore the Fintra-AI platform and its core features through the following screenshots.
From expense tracking and financial analytics to AI-powered insights, Fintra-AI provides an intelligent and modern approach to personal finance management.
Turn financial data into intelligent, personalized, and actionable financial decisions.
Traditional expense trackers primarily answer:
"Where did my money go?"
Fintra-AI aims to answer much more:
- Where am I spending the most?
- What spending patterns are emerging?
- Am I likely to exceed my budget?
- How healthy is my financial behavior?
- What can I do to improve my savings?
- Which financial goals are achievable?
- Can AI help me plan my next month?
- Can machine learning identify unusual financial activity?
This creates a shift from:
Expense Tracking → Financial Intelligence
- Expense Tracking
- Income Tracking
- Budget Management
- Savings Goals
- Financial Dashboard
- Monthly Reports
- Cash Flow Analysis
- Category-Based Expenses
- Transaction Management
- Financial Summary
Powered by Google Gemini AI, Fintra-AI can provide intelligent financial assistance such as:
- AI Financial Advisor
- AI Budget Suggestions
- AI Expense Insights
- AI Goal Planning
- AI Spending Analysis
- AI Financial Chat Assistant
- Personalized Recommendations
- Financial Data Summarization
- Smart Budget Recommendations
The planned ML engine focuses on transforming historical financial data into predictive insights.
| Capability | Purpose |
|---|---|
| 📂 Expense Classification | Predict appropriate expense categories |
| 📈 Spending Forecasting | Estimate future spending |
| 💰 Budget Prediction | Predict potential budget overruns |
| 🚨 Fraud Detection | Identify suspicious financial patterns |
| ❤️ Financial Health Score | Estimate overall financial behavior |
| 🎯 Goal Prediction | Estimate goal completion probability |
| 💡 Recommendation Engine | Generate personalized financial suggestions |
| 📊 Trend Detection | Identify changes in spending behavior |
Fintra-AI provides a data-driven view of personal finances.
- Expense Analytics
- Income Analytics
- Category Distribution
- Monthly Spending Trends
- Cash Flow Analysis
- Budget Utilization
- Savings Analysis
- Financial Reports
- Interactive Charts
- Historical Comparisons
Financial Transactions
│
▼
Data Processing
│
▼
┌───────────────────┐
│ Feature Extraction│
└─────────┬─────────┘
│
▼
ML / AI Engine
│
┌─────┴─────┐
▼ ▼
Prediction Insights
│ │
└─────┬─────┘
▼
Personalized
Recommendations
Security is a core component of Fintra-AI because financial data is highly sensitive.
- 🔑 Clerk Authentication
- 🛡️ Arcjet Security
- 🔒 Protected APIs
- 👤 Session Management
- 🗄️ Secure Database Access
- 🎭 Role-Based Access Control
- 🔐 Environment-Based Secrets
- 🚫 No sensitive credentials committed to source control
Never commit
.env, API keys, database credentials, or authentication secrets to Git.
Fintra-AI can provide timely financial notifications including:
- 📩 Email Notifications
⚠️ Budget Alerts- 🎯 Savings Goal Reminders
- 📊 Monthly Financial Reports
- 🔔 Financial Activity Notifications
Email infrastructure is powered by Resend.
- Next.js
- React
- Tailwind CSS
- Shadcn UI
- Next.js Server Actions
- Prisma ORM
- PostgreSQL
- Google Gemini API
- Scikit-Learn
- TensorFlow
- PyTorch
- XGBoost
- Prophet
- Clerk
- Resend
- Arcjet
- Vercel
┌──────────────────────┐
│ User / Client │
└──────────┬───────────┘
│
▼
┌──────────────────────┐
│ Next.js App │
│ UI + Server Logic │
└──────────┬───────────┘
│
┌──────────────┼──────────────┐
│ │ │
▼ ▼ ▼
┌───────────┐ ┌───────────┐ ┌────────────┐
│ Clerk │ │ Arcjet │ │ Resend │
│ Auth │ │ Security │ │ Email │
└───────────┘ └───────────┘ └────────────┘
│
▼
┌──────────────────┐
│ Prisma │
│ ORM │
└────────┬─────────┘
│
▼
┌──────────────────┐
│ PostgreSQL │
│ Database │
└────────┬─────────┘
│
▼
┌─────────────────────────┐
│ AI / ML Layer │
├─────────────────────────┤
│ Google Gemini │
│ ML Prediction Models │
│ Recommendation Engine │
│ Financial Analytics │
└─────────────────────────┘
Fintra-AI/
│
├── ai-finance-platform/ # Next.js Frontend Application
│ ├── app/ # App Router & Pages
│ ├── actions/ # Server Actions
│ ├── components/ # Reusable UI Components
│ ├── hooks/ # Custom React Hooks
│ ├── lib/ # Utilities & Configurations
│ ├── public/ # Static Assets
│ ├── emails/ # Email Templates
│ ├── middleware.js # Authentication & Middleware
│ ├── next.config.mjs # Next.js Configuration
│ ├── package.json
│ └── ...
│
├── backend/ # FastAPI Backend
│ ├── app/
│ │ ├── api/
│ │ │ └── v1/ # Versioned REST API
│ │ │ ├── transactions.py # Transaction Management
│ │ │ ├── budgets.py # Budget Management
│ │ │ ├── analytics.py # Financial Analytics
│ │ │ ├── predictions.py # ML Predictions
│ │ │ ├── ai.py # AI-powered Features
│ │ │ └── reports.py # Financial Reports
│ │ │
│ │ ├── core/ # Core Backend Configuration
│ │ │ ├── config.py # Application Configuration
│ │ │ ├── database.py # Database Configuration
│ │ │ └── security.py # Authentication & Security
│ │ │
│ │ ├── models/ # Database Models
│ │ ├── schemas/ # Pydantic Schemas
│ │ │
│ │ ├── services/ # Business Logic Layer
│ │ │ ├── transaction_service.py
│ │ │ ├── analytics_service.py
│ │ │ ├── prediction_service.py
│ │ │ └── ai_service.py
│ │ │
│ │ └── main.py # FastAPI Application Entry Point
│ │
│ ├── tests/ # Backend Tests
│ ├── requirements.txt # Python Dependencies
│ └── Dockerfile # Backend Container
│
├── ml/ # Machine Learning Pipeline
│ ├── datasets/ # ML Datasets
│ ├── preprocessing/ # Data Preprocessing
│ ├── features/ # Feature Engineering
│ ├── training/ # Model Training
│ │ ├── train_categorization.py # Expense Categorization
│ │ ├── train_forecasting.py # Financial Forecasting
│ │ └── train_anomaly.py # Anomaly Detection
│ │
│ ├── inference/ # Model Inference
│ ├── evaluation/ # Model Evaluation
│ └── models/ # Trained ML Models
│
├── data_pipeline/ # Data Engineering Pipeline
│ ├── ingestion/ # Data Ingestion
│ ├── preprocessing/ # Data Cleaning & Processing
│ └── jobs/ # Automated Data Jobs
│
├── notebooks/ # Jupyter Notebooks
│ ├── 01_eda.ipynb # Exploratory Data Analysis
│ ├── 02_categorization.ipynb # Expense Categorization
│ └── 03_forecasting.ipynb # Financial Forecasting
│
├── infrastructure/ # DevOps & Cloud Infrastructure
│ ├── docker/ # Docker Configuration
│ ├── monitoring/ # Monitoring & Observability
│ └── terraform/ # Infrastructure as Code
│
├── .github/
│ └── workflows/ # CI/CD Pipelines
│ ├── frontend-ci.yml # Frontend CI
│ ├── backend-ci.yml # Backend CI
│ └── ml-ci.yml # ML CI
│
├── docs/ # Project Documentation
├── Screenshots/ # Application Screenshots
│
├── docker-compose.yml # Multi-Service Docker Setup
├── .gitignore # Git Ignore Rules
└── README.md # Project Documentation
Current implementation note: The runnable product is the Next.js/Prisma application in
ai-finance-platform/. Theml/directory contains separate Python research and inference pipelines. The FastAPI backend, data-pipeline jobs, Terraform/monitoring infrastructure, and CI services shown in the roadmap are planned surfaces and are not currently implemented as production services.
Fintra-AI currently uses a Next.js application with server actions and a separate ML research layer. The application integrates Clerk, Prisma/PostgreSQL, Gemini-assisted receipt/report features, Resend email, and Inngest jobs. Future backend, data-engineering, and MLOps layers should be added only when their source code and deployment instructions are committed.
┌─────────────────────────┐
│ User / Client │
└────────────┬────────────┘
│
▼
┌─────────────────────────┐
│ Next.js Frontend │
│ ai-finance-platform/ │
└────────────┬────────────┘
│ REST API
▼
┌─────────────────────────┐
│ FastAPI Backend │
│ backend/ │
└───────┬─────────┬───────┘
│ │
┌────────────┘ └─────────────┐
▼ ▼
┌──────────────────┐ ┌──────────────────┐
│ ML Services │ │ Data Pipeline │
│ ml/ │ │ data_pipeline/ │
└────────┬─────────┘ └────────┬─────────┘
│ │
▼ ▼
┌──────────────────┐ ┌──────────────────┐
│ Trained Models │ │ Data Processing │
│ Forecasting │ │ Ingestion │
│ Categorization │ │ ETL Jobs │
│ Anomaly Detection│ └──────────────────┘
└──────────────────┘
│
▼
┌──────────────────┐
│ Database / Data │
│ Layer │
└──────────────────┘
| Layer | Directory | Responsibility |
|---|---|---|
| Frontend | ai-finance-platform/ |
User interface, dashboards and client interactions |
| Backend | backend/ |
REST APIs, authentication and business logic |
| AI/ML | ml/ |
Training, inference and financial intelligence |
| Data Engineering | data_pipeline/ |
Data ingestion, preprocessing and automation |
| Research | notebooks/ |
EDA, experimentation and model research |
| Infrastructure | infrastructure/ |
Docker, monitoring and cloud deployment |
| CI/CD | .github/workflows/ |
Automated testing and deployment |
| Documentation | docs/ |
Technical and project documentation |
Fintra-AI's ML layer is designed around three primary intelligence modules:
- Expense Categorization — Automatically classifies financial transactions.
- Financial Forecasting — Predicts future spending and financial trends.
- Anomaly Detection — Identifies unusual or potentially suspicious transactions.
This separation keeps model development and experimentation independent from the production FastAPI services, making the platform easier to scale, test, deploy, and maintain.
---
# 🚀 Getting Started
Follow these steps to run Fintra-AI locally.
## 1️⃣ Clone the Repository
```bash
git clone https://github.com/Ashwinchauhan89/Fintra-AI.git
cd Fintra-AI
The runnable application is located in ai-finance-platform/. Install its dependencies from that directory:
cd ai-finance-platform
npm ciCreate ai-finance-platform/.env.local from the committed ai-finance-platform/.env.example template.
DATABASE_URL=
DIRECT_URL=
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY=
CLERK_SECRET_KEY=
NEXT_PUBLIC_CLERK_SIGN_IN_URL=/sign-in
NEXT_PUBLIC_CLERK_SIGN_UP_URL=/sign-up
NEXT_PUBLIC_CLERK_AFTER_SIGN_IN_URL=/onboarding
NEXT_PUBLIC_CLERK_AFTER_SIGN_UP_URL=/onboarding
GEMINI_API_KEY=
RESEND_API_KEY=
ARCJET_KEY=
⚠️ Keep all secrets private and never commit.envto Git.
npx prisma generatenpx prisma db pushnpm run devThe application will be available at:
http://localhost:3000
Fintra-AI uses Prisma ORM with PostgreSQL.
npx prisma generatenpx prisma db pushnpx prisma studioPrisma Studio can be used to inspect and manage development database records.
Recommended development flow:
Create Issue
│
▼
Create Feature Branch
│
▼
Implement Feature
│
▼
Run Tests / Checks
│
▼
Review Changes
│
▼
Commit
│
▼
Push Branch
│
▼
Open Pull Request
│
▼
Code Review
│
▼
Merge
Recommended naming convention:
feature/expense-analytics
feature/ai-budget-advisor
feature/ml-spending-prediction
fix/authentication-error
fix/dashboard-loading
docs/update-readme
docs/ml-documentation
| Document | Description |
|---|---|
README.MD |
Project overview and setup |
CONTRIBUTION.MD |
Contribution guidelines |
PLANNING.MD |
Development roadmap and milestones |
MACHINELearning.md |
ML architecture and pipeline |
ai-finance-platform/README.md |
Frontend environment variables |
ml/README.md |
ML pipeline setup and usage |
- Authentication
- Dashboard
- Expense Management
- Income Management
- Advanced Transaction Management
- Budget Planning
- Reports
- Analytics
- Advanced Cash Flow Analytics
- Financial Trend Detection
- AI Financial Advisor
- Gemini Integration (receipt scanning and monthly reports)
- AI Budget Recommendations
- Personalized Financial Insights
- AI Financial Copilot
- Expense Classification
- Spending Prediction
- Budget Forecasting
- Fraud Detection
- Financial Health Score
- Goal Completion Prediction
- Recommendation Engine
- OCR Receipt Scanner (Gemini-assisted)
- Investment Tracker
- Family Wallet
- Voice Finance Assistant
- PWA Support
- Multi-Currency Support
- Open Banking Integration
- Explainable AI
- MLOps Pipeline
📄 See PLANNING.MD for the detailed roadmap.
Fintra-AI is designed as an open-source collaboration project and welcomes contributors interested in software engineering, Artificial Intelligence, Machine Learning, cybersecurity, data analytics, and product development.
| Level | Suitable Contributions |
|---|---|
| 🟢 Beginner | Documentation, UI improvements, testing, bug fixes |
| 🟡 Intermediate | APIs, database features, analytics, backend modules |
| 🔴 Advanced | AI, ML, fraud detection, predictive analytics, OCR |
Frontend
├── Dashboard
├── Charts
└── UI/UX
Backend
├── APIs
├── Database
└── Server Actions
AI
├── Gemini Integration
├── Financial Advisor
└── Recommendations
ML
├── Classification
├── Forecasting
├── Fraud Detection
└── Financial Health
DevOps
├── CI/CD
├── Testing
└── Deployment
Contributions are welcome from:
- 👨💻 Software Developers
- 🎨 UI/UX Designers
- 🤖 AI Engineers
- 🧠 Machine Learning Engineers
- 📊 Data Scientists
- 🔐 Cybersecurity Developers
- 🧪 QA/Test Engineers
- 📝 Technical Writers
- Fork the repository.
- Create a feature branch.
- Implement your changes.
- Test your changes locally.
- Commit your changes.
- Push your branch.
- Open a Pull Request.
- Participate in code review.
Before contributing, please read:
CONTRIBUTION.MD
Fintra-AI aims to evolve from a traditional finance tracker into a complete AI Financial Intelligence Platform.
┌──────────────────────┐
│ FINTRA-AI │
│ Financial Intelligence│
└──────────┬───────────┘
│
┌────────────────────┼────────────────────┐
│ │ │
▼ ▼ ▼
AI Advisor ML Prediction Analytics
│ │ │
▼ ▼ ▼
Personalized Future Spending Financial
Recommendations Forecasting Insights
│ │ │
└────────────────────┼────────────────────┘
▼
Smarter Financial Decisions
- 🤖 AI Financial Copilot
- 🧠 Advanced ML Prediction Engine
- 🚨 Real-Time Fraud Detection
- 🏦 Open Banking Integration
- 🎙️ Voice Finance Assistant
- 📷 Smart Receipt OCR
- 📈 Investment Portfolio Management
- 🌍 Multi-Currency Support
- 🔍 Explainable AI
- ⚙️ Production MLOps Pipeline
- 👨👩👧 Family Finance Management
- 📱 Progressive Web Application
This project is licensed under the MIT License.
See the LICENSE file for complete license information.
Project Lead — Fintra-AI
GitHub: @Ashwinchauhan89
If you find Fintra-AI useful, interesting, or valuable for learning:
Every contribution helps improve the project and grow the open-source community around Fintra-AI.