Evaluating LLM features before you launch them
Shipping an AI feature without an evaluation harness is shipping a system nobody can prove works. Here is the eval setup we build on every engagement.
This article is a placeholder body. Connect a headless CMS — Sanity or Contentful are both a good fit for this setup — and replace this block with the rendered portable-text or rich-text field. The route, metadata, breadcrumb and BlogPosting structured data are already wired up, so publishing real content is a data change rather than a code change.
Suggested structure
Keep articles to a single H1 (the title above) with H2s for major sections and H3s beneath them. That hierarchy is what search engines parse for featured snippets, and it is also what screen-reader users navigate by.
Wiring up the CMS
Swap the static posts array in src/data/content.ts for a CMS query, keep generateStaticParams pointing at the same source, and the whole blog becomes statically generated with on-demand revalidation.