22. Glossary
| Term | Definition |
|---|---|
| DX Platform | Developer Experience Platform — Vercel's self-description; infrastructure + tooling optimised for developer productivity |
| Edge Runtime | V8 isolate-based runtime for ultra-fast, globally distributed code execution at CDN PoPs |
| Fluid Compute | Vercel's compute model that allows function instances to serve multiple requests concurrently, with Active CPU pricing |
| Active CPU pricing | Billing model that charges only for actual CPU execution time, not I/O wait time |
| ISR | Incremental Static Regeneration — static pages that regenerate in the background after a time interval or webhook trigger |
| RSC | React Server Components — server-rendered components that ship zero JavaScript to the client |
| PoP | Point of Presence — a CDN server location close to end users |
| Stale-while-revalidate | Cache strategy that serves stale content immediately while refreshing in the background |
| Cache stampede | When many concurrent requests all trigger cache regeneration simultaneously — Vercel prevents with request collapsing |
| Skew protection | Enterprise feature ensuring users always receive responses from the same deployment version that served their initial HTML |
| Edge Config | Globally replicated key-value store for ultra-low latency reads at the edge |
| v0 | Vercel's AI-powered frontend app builder that generates Next.js code from natural language |
| AI SDK | Vercel's open-source SDK for building AI-powered applications with unified provider interface |
| Turbopack | Vercel's Rust-based bundler, successor to webpack, built into Next.js |
| TTFB | Time to First Byte — the time from request to first byte of the response |
| LCP | Largest Contentful Paint — Core Web Vital measuring when the largest visible element renders |
| CLS | Cumulative Layout Shift — Core Web Vital measuring visual stability |
| INP | Interaction to Next Paint — Core Web Vital measuring responsiveness to user input |
| Draft Mode | Next.js feature that bypasses ISR cache to preview unpublished CMS content |
| Request collapsing | Vercel CDN feature that coalesces multiple concurrent requests for the same uncached resource into one origin call |
| Composable architecture | Architectural pattern assembling best-of-breed services (CMS, commerce, auth, search) connected through APIs |
| Headless CMS | Content management system that provides content via API rather than rendering its own frontend |
| MACH | Architecture principle: Microservices, API-first, Cloud-native, Headless — the framework many Vercel enterprise customers use |
| use-workflow | Vercel's open-source library for converting functions into durable, retry-able workflows |
| Bytecode caching | V8 feature that caches compiled bytecode across Fluid Compute invocations, eliminating parse/compile overhead |
Study tip: After reading each section, go to vercel.com/docs and find the documentation for that concept. Then deploy something that uses it. You will be expected to demo and discuss this in a real customer context.
Key resources:
- Vercel Docs: vercel.com/docs
- Next.js Docs: nextjs.org/docs
- Vercel Blog (Engineering): vercel.com/blog/category/engineering
- Vercel Ship AI 2025 talks: vercel.com/ship-ai
- AI SDK Docs: sdk.vercel.ai