I'm a software engineer who got curious about whether the AI infrastructure buildout follows the same patterns as the telecom and dot-com cycles. So I built a composite stress index that tracks 12 financial signals: credit spreads, capex/EBITDA ratios, NVIDIA revenue deceleration, neocloud leverage, hyperscaler debt issuance - and maps them against historical benchmarks.

The data comes from SEC filings, FINRA TRACE, and public market data. Current reading: the AI buildout's capex intensity is tracking closer to AT&T's telecom peak (72% of EBITDA) than the dot-com equity bubble.

Stack: Go backend, Fly.io deployment (collector, private API and public API deployments separate), Postgres database. Next.js frontend deployed on Vercel.

Would love feedback on the methodology and signal weighting. The full breakdown is on the website.