AI Supply Chain Monitor

What it is

A Next.js 16 dashboard that presents a researched 14-layer map of the AI supply chain, from raw materials to deployed inference, as an editorial ETF-style basket of about 85 public companies. Each company gets fundamentals, filings and news sentiment, and a composite score blended with Claude summaries. It renders from a static seed with no keys and upgrades to Supabase-backed live data when configured.

The problem

Investors lack a structured view of where the real bottlenecks and dependencies sit across the AI hardware and power stack.

How it works

1. A dual data path: store.ts reads a Supabase scores view and falls back to a static seed plus deterministic mock scores, so the UI always renders.
2. Ingestion is browser-orchestrated from /admin, firing one API request per ticker at 3 to 5 in parallel so each Vercel function stays under its timeout, with no cron or workers.
3. Pulls fundamentals and prices from Financial Modeling Prep, filings from SEC EDGAR and news feeds, and uses Claude for scoring and explainers.
4. Pages cover layers, companies, rankings, ETF performance, a primer and circular-flow analysis.

Stack

Disciplines

For your business

Research dashboards that combine licensed market data, filings and LLM summaries into one scored, browsable view.