How vishveswara31/finance-ai works
FinanceAI is a Flask finance assistant with four user-facing workflows: expense tracking, 50/30/20 budget generation, transaction fraud scoring, and company investment recommendations. A browser submits form data to Flask routes; each route validates input and invokes its corresponding agent, which returns a result and, where applicable, persists records in SQLite. Investment lookup also reads company reference data. The source excerpts establish these calls and data accesses; the HTML template and styling exist in the tree but were not sampled, so their internal behavior is not expanded here.
Main entry points: app.py API routes · database.py SQLite access · investment_agent.py Investment lookup · script.js Form client
Components 11
Browser interface
- Finance page
templates/index.html - Form client
static/script.js
Flask API
- API routes
app.py
Finance workflows
- Expense tracker
agents/expense_agent.py - Budget planner
agents/budget_agent.py - Fraud scorer
agents/fraud_agent.py - Investment lookup
agents/investment_agent.py
Data stores
- SQLite access
database.py - Finance database
database - Company catalog
data/companies.json
Other components
- Finance user
Connections 14
- Finance user → Finance page: uses
- API routes → Finance page: renders
app.py - Form client → API routes: posts forms
static/script.js - API routes → Form client: returns JSON
app.py - API routes → SQLite access: initializes
app.py - API routes → Expense tracker: invokes
app.py - API routes → Budget planner: invokes
app.py - API routes → Fraud scorer: invokes
app.py - API routes → Investment lookup: invokes
app.py - Expense tracker → SQLite access: uses
agents/expense_agent.py - Budget planner → SQLite access: uses
agents/budget_agent.py - Investment lookup → SQLite access: uses
agents/investment_agent.py - SQLite access → Finance database: connects
database.py - Investment lookup → Company catalog: reads catalog
agents/investment_agent.py