spending-analysis

🛠️ Prototype for the Hacknovate Hackathon 2025 (Agentic AI Track). UPISensei: An autonomous personal finance agent that understands, analyses, and explains UPI transactions automatically, transforming raw bank statement data into actionable financial insights using agentic AI. 💵

Personal finance PWA — CSV transaction import, auto-categorization, spending visualizations, AI-powered insights

CLI pipeline to parse PDF bank statements, auto-categorise transactions, and generate spending analysis.

Analysis of debt levels and public spending quality: Brazil vs. LATAM, using IMF data extracted via API.

Offline UPI Spending Analysis and Tracking by SMS and Notification

Análise exploratória dos gastos da CEAPS do Senado Federal em 2022, identificando padrões de distribuição, sazonalidade, outliers e principais tipos de despesa com dados públicos.

Data-driven analysis of payment methods and their impact on spending awareness and financial management.

Financial spending behavior analysis and dashboard visualization identifying budgeting patterns, merchant concentration, and seasonal spending trends.

Only wallet you need.

A TypeScript CLI that retrieves the previous day's **initial prime federal contract awards** from USAspending, identifies recipients associated with companies listed on Nasdaq, the NYSE, or NYSE American, groups the results into one row per ticker

Fuelsubsidy landing page, an app that helps vehicle owners save on fuel and track their usage. It highlights key features like digital vouchers for discounts, rewards for purchases and referrals, and tools for monitoring fuel spending, trips, and driving habits.

End-to-end compliance audit pipeline validating 12,860 transactions and surfacing $731K in financial exposure across 6 departments.

AI-powered financial behaviour intelligence system built for FINNMIN Hackathon 2026 (PS3). Detects unnecessary spending patterns from transaction data, simulates counterfactual savings, and delivers personalized, explainable cost-saving recommendations via an AI financial coach — no black-box AI, just structured behavior detectors + LLM narration.