What I built
Global Solo is a structural risk diagnostic platform for solo founders operating across borders. Users answer 18–45 questions about their business structure, and the system generates a META Risk Profile mapping risk across Money, Entity, Tax, and Accountability dimensions.
Free risk screening tool: globalsolo.global/tools/risk-check
How Vercel + Next.js powers it
- ISR + static generation for 49 blog articles and 7 case studies — knowledge base pages are statically generated with daily revalidation, giving near-instant page loads for SEO-critical content
- API routes handle a multi-stage LLM pipeline — three stages (analyzer → narrator → assembler) run in Vercel serverless functions with 60-90s timeouts for premium reports
- Edge middleware for auth session refresh — Supabase Auth sessions are refreshed via Next.js middleware on every request
- Stripe webhooks + Vercel Cron — payment processing and scheduled email sequences (drip campaigns, compliance reminders) all run on Vercel infrastructure
- Dynamic OG images — generated per-page using
opengraph-image.tsxfor social sharing
Stack
Next.js 16, React 19, Tailwind CSS 4, Supabase (PostgreSQL + Auth), Stripe, OpenAI + Anthropic (dual LLM provider), Resend, Sentry, Vercel
Solo founder story
I built Global Solo because I’ve lived cross-border complexity firsthand — running an established consumer brand (AirPop) across US/China markets while building new ventures. The entire platform was shipped in ~6 weeks as a solo technical founder. Vercel’s DX — from local dev to production deploy with zero DevOps — made that timeline possible.
The diagnostic engine processes questionnaire responses through a deterministic signal mapper, scoring engine, and LLM-enhanced narrative layer. Same input always produces the same structural analysis. The LLM controls wording, not scores.