FinRisk AI - Smart Financial Risk Detector

Dynamic category breakdown and telemetry visualization populated via sample student data sandbox
FinRisk AI Dashboard: Real-time financial health score, 30-days velocity forecast, and expense metrics
Automated outlier detection: High Anomaly risk alert triggered instantly on spending threshold breach
one-click audit pipeline: Automated CSV financial health report generation for external ledger sync.

Inspiration

Traditional personal finance applications act merely as passive ledgers—users manually log expenditures only to realize days later that their monthly budget has been depleted. We wanted to build an active, real-time guardian: an engine that monitors spending velocity, detects abnormal spikes instantly, and acts as an early warning system before financial distress occurs.

What it does

FinRisk AI is an intelligent personal financial monitoring engine:

Real-Time Outlier & Anomaly Detection: Instantly scores incoming expenses against baseline distributions, immediately flagging abnormal spikes with dynamic risk indicators (NORMAL vs. HIGH ANOMALY) and visual warning banners.
Predictive 30-Day Expense Velocity: Tracks current burn rates to forecast month-end variance against allocated budget thresholds.
Spending Distribution Breakdown: Visualizes capital outflow across key categories (Food, Shopping, Transport, Entertainment, Utilities).
Automated Financial Coach: Delivers real-time, context-aware financial coaching to curb impulse purchases based on category concentration.
Audit & Export Pipeline: Ingests local transaction logs and generates one-click CSV financial audit reports for spreadsheet reconciliation.
Telemetry Sandbox: Includes a built-in "Load Sample Student Data" trigger for rapid benchmarking.

How we built it

Core Engine & Architecture: Built with semantic HTML5, modern CSS3 (fintech dark-mode dashboard), and modular Vanilla JavaScript (ES6+).
Client-Side Analytics: Dynamic mathematical forecasting algorithms, lightweight SVG/Canvas data visualization, and an in-browser CSV parsing/generation pipeline.
AWS Integration Roadmap: Designed for seamless integration with AWS EventBridge to handle event-driven transaction streams and Amazon Bedrock for generative personal financial roadmaps.

Challenges we ran into

Designing a zero-dependency anomaly detection and forecasting model entirely in native client-side JavaScript that updates instantaneously without UI lag.
Structuring mathematical risk weights so single large transactions accurately adjust health scores without corrupting historical trendlines.

Accomplishments that we're proud of

Delivering a lightweight, responsive fintech interface with zero build tools or heavy framework overhead.
Implementing an instant anomaly alert system and automated CSV audit report generator in a self-contained web architecture.

What we learned

How to structure clean, event-driven vanilla JavaScript architectures.
Practical techniques for calculating expenditure velocity and modeling proactive personal financial health metrics.

What's next for FinRisk AI

Integrating AWS Lambda and Amazon Bedrock to provide multi-modal bank statement analysis and conversational AI advisory.
Multi-currency support and real-time open banking API webhooks.