GPU-accelerated cash-flow intelligence dashboard for freelancers and small business owners.
�� Live demo: https://cash-flow-radar-avg53pwdk4mxpmt6rev5tz.streamlit.app/
CashFlow Radar ingests transaction data, cleans and aggregates it using NVIDIA RAPIDS (cuDF), detects spending anomalies, and forecasts the next 30 days of cash position — surfacing a single Cash-Flow Risk Score and ranked alerts so users can decide, in seconds, whether a purchase or expense is safe to make.
Freelancers and small business owners often only discover a cash-flow problem after it's already happened. This tool answers the question that actually matters: "Can I afford this, and what should I do about my finances right now?"
- Data processing: NVIDIA RAPIDS, cuDF (GPU-accelerated cleaning/aggregation)
- Modeling: cuML-based anomaly detection, lightweight time-series forecasting
- Dashboard: Streamlit + Plotly
- No GCP dependency — built entirely on the NVIDIA acceleration layer
- 📊 Real-time dashboard with balance, spend trends, and risk score
- 🔮 30-day cash-flow forecast with confidence bands
- 🚨 Anomaly detection for unusual/suspicious transactions
- ⚡ Live CPU (pandas) vs GPU (cuDF) benchmark, demonstrating measurable acceleration
- 🧭 Full navigation: Dashboard, Accounts, Transactions, Analytics, Budgets, Settings
```bash pip install -r requirements.txt streamlit run app.py ```
Data is generated and processed in a Colab notebook (GPU-accelerated via RAPIDS), then exported as CSV/JSON files consumed by this dashboard. See `/data` for sample output.
Built for the GenAI APAC Edition Hackathon.