How nareshjo001/expense-manager works
BALENISA is a full-stack personal-finance platform: users record and import financial activity, manage budgets and recurring expenses, then explore charts, insights, and grounded SIA answers. The React client calls an Express API, which owns authorization, financial writes, analytics, and external-service calls. Major documented branches include receipt OCR, exports, notifications, and category prediction/retraining. Some subsystem links are documented only at a high level; unsampled implementation details are not inferred.
Main entry points: app.js Express API · Schemas.js MongoDB · api.routes.js API routes · semanticPipeline.js Grounded query pipeline
Components 24
React client
- Finance UI
frontend/src/App.js - API clients
frontend/src/api/axios.js
Finance API
- Authentication
backend/Controllers/AuthControllers/login.js - Express API
backend/app.js - API routes
backend/Routes/api.routes.js - Expenses and income
backend/Controllers/ExpenseControllers/addexpense.js - Budgets
backend/Controllers/BudgetControllers/setbudget.js - Recurring expenses
backend/Controllers/RecurringExpenses/recurring.js - Import sessions
backend/Controllers/ImportControllers/createSession.js - Data exports
backend/Controllers/Export/create.js - Receipt OCR and inbox
backend/Services/ReceiptServices/receiptIngestService.js
Analytics and SIA
- Charts
backend/Services/ChartServices/chart.service.js - Financial insights
backend/analytics/reportGenerator.js - SIA assistant
backend/Controllers/SiaControllers/ask.js - Grounded query pipeline
backend/sia/semanticPipeline.js
Platform services
- MongoDB
backend/config/Schemas.js - Redis cache
backend/config/redis.js - ML service client
backend/utils/mlServiceClient.js
ML service
- Category prediction
ml-service/app.py - Live model manager
ml-service/inference/predictor_manager.py - Model retraining
ml-service/training/retrain_pipeline.py
Other components
- User
- LLM provider
- Firebase push
Connections 31
- User → Finance UI: uses
README.md - Finance UI → API clients: requests
README.md - API clients → Express API: REST calls
README.md - Express API → Authentication: mounts routes
backend/app.js - Express API → API routes: mounts
backend/app.js - API routes → Budgets: routes to
backend/Routes/api.routes.js - API routes → Recurring expenses: routes to
backend/Routes/api.routes.js - API routes → Import sessions: routes to
backend/Routes/api.routes.js - API routes → Data exports: routes to
backend/Routes/api.routes.js - Express API → Expenses and income: serves routes
backend/app.js - Express API → Receipt OCR and inbox: serves routes
backend/app.js - Express API → Charts: serves routes
backend/app.js - Express API → Financial insights: serves routes
backend/app.js - Express API → SIA assistant: serves routes
backend/app.js - Expenses and income → MongoDB: writes data
README.md - Budgets → MongoDB: stores budgets
README.md - Recurring expenses → MongoDB: stores definitions
README.md - Import sessions → MongoDB: stores sessions
- Data exports → MongoDB: reads finance data
- Receipt OCR and inbox → MongoDB: persists receipts
backend/Services/ReceiptServices/receiptIngestService.js - Charts → MongoDB: reads activity
README.md - Financial insights → MongoDB: reads activity
README.md - SIA assistant → Grounded query pipeline: uses pipeline
backend/sia/semanticPipeline.js - Grounded query pipeline → MongoDB: queries user facts
- Grounded query pipeline → LLM provider: optional explanation
README.md - Express API → Redis cache: uses cache
README.md - Express API → ML service client: requests predictions
backend/utils/mlServiceClient.js - ML service client → Category prediction: service calls
README.md - Category prediction → Live model manager: loads predictor
ml-service/app.py - Category prediction → Model retraining: starts retraining
ml-service/app.py - Express API → Firebase push: sends notifications
README.md