FinanceWeb / Design
Portfolio Scatter
A deployed full-stack risk analyzer that computes VaR, Sharpe, beta, and sector look-through, with an AI analyst and PDF or image import.
- 15+
- risk metrics
- 4
- import formats
- AI
- PDF + image parse
- live
- web app
Demo
Portfolio Scatter · demo
00:00 / 00:25
Overview
A live, full-stack portfolio risk analyzer. Holdings come in as CSV, XLSX, PDF, or even a screenshot: structured files parse deterministically, while PDFs and images fall back to an AI parser. ETFs are expanded via sector and single-stock look-through1, and a risk engine computes a full metric suite (VaR2, Sharpe, Sortino, beta, drawdown, concentration, and stress scenarios), rendered as interactive dashboards alongside an on-demand AI analyst. Accounts and saved portfolios sit behind real authentication, and the AI prompts are regression-tested3 before release.
- 1look-through · Expanding each ETF into the stocks it holds, so sector weights are measured on the underlying companies rather than the fund wrapper.
- 2VaR · Value at Risk: the loss a portfolio is not expected to exceed on a normal day at a stated confidence, e.g. 95%.
- 3regression-tested · Every prompt has fixed inputs with expected outputs; a change that alters an output fails the run before it ships.
Highlights
- 15+ risk metrics: VaR, Sharpe, Sortino, beta, drawdown, concentration, and stress scenarios
- Format-agnostic import: CSV and XLSX parse directly, with an AI fallback for PDFs and screenshots
- ETF look-through to underlying sector and single-stock exposure
- Interactive risk dashboards paired with an on-demand AI analyst
- Accounts and saved portfolios, with the AI prompts regression-tested before release
Stack & tools
- Next.js 16
- React 19
- TypeScript
- shadcn/ui
- Tailwind 4
- Recharts
- Zustand
- Upstash Redis
- Playwright
- Vitest