preissersolutions.com — 164 Pages, Custom Next.js, Built for AI Search
A bespoke marketing and AI-search platform for Preisser Solutions itself — Next.js 15, an 8,800-line custom design system, AI-native infrastructure, and 164 statically-rendered pages.
Generic agency-template marketing pages that did not differentiate.
Before this build, the Preisser Solutions web surface was the same kind of generic agency template that every consultancy ends up with — pages that look like every other page, no AI-search infrastructure, no honest case-study layer, and no way for AI agents to cite the work cleanly.
For a consultancy whose pitch is building production-grade custom software, that surface was actively undermining the offer. The rebuild treated the site as a first-class engineering artifact and as the company's most public proof of capability.
A full custom build — design system, components, and AI-native infrastructure.
Preisser Solutions built the site end-to-end on Next.js 15, React 19, and TypeScript, with Tailwind v4 layered over an 8,800-line custom design system that uses CSS cascade layers, HDS tokens, and bespoke component utilities. GSAP 3.12.7 (single registration through `lib/gsap.ts`) and Framer Motion 12 handle animations. The output ships as a static export to Cloudflare Pages with edge middleware for canonicalization.
Two component systems carry the design language. The CaseStudyPage component (714 lines) drives every case-study route with a radial-glow hero, glass metric chips, gradient-text numbers, and GSAP scroll-stagger reveals. The LocationPage component drives the geographic surface with a city-led hero, nearby-towns chip grid, service cards, and a process timeline.
AI-native infrastructure is layered on top of the marketing surface. `llms.txt` and `llms-full.txt` are served at the edge for AI agents. A JSON-LD schema graph wires Organization, Person, LocalBusiness, WebPage, BreadcrumbList, FAQPage, and Article entities together. Agent discovery endpoints live at `/.well-known/agent-card.json` and `/.well-known/mcp/server-card.json`. Content-signal headers (`ai-train=yes, search=yes, ai-input=yes`) and a fully open X-Robots-Tag complete the citation surface.
Stack, scale, and AI-native surface.
Next.js 15 + React 19 + TypeScript with strict mode
Tailwind v4 over an 8,800-line custom design system (CSS cascade layers, HDS tokens)
GSAP 3.12.7 (single registration) + Framer Motion 12 for animation
Static export to Cloudflare Pages with edge middleware canonicalization
~164 statically-rendered URLs in sitemap across services, case studies, locations, industries, blog, comparisons, insights, and use-cases
AI-native infrastructure: llms.txt, llms-full.txt, agent discovery endpoints, JSON-LD schema graph
Component systems
- CaseStudyPage (714 lines) — radial-glow hero, glass metric chips, gradient-text numbers, GSAP scroll-stagger
- LocationPage — geographic hero, nearby-towns chip grid, service cards, process timeline
- Bespoke design tokens for color, motion, and spacing — no template libraries
- Cascade-layered CSS architecture for predictable specificity
AI-native infrastructure
- llms.txt and llms-full.txt served at the edge for AI agents
- JSON-LD schema graph (Organization, Person, LocalBusiness, WebPage, BreadcrumbList, FAQPage, Article)
- Agent discovery endpoints at /.well-known/agent-card.json and /.well-known/mcp/server-card.json
- Content-signal headers: ai-train=yes, search=yes, ai-input=yes
- X-Robots-Tag fully open for AI search
Content surface
- ~164 statically-rendered URLs at build time
- 21+ case-study detail pages with consistent data contract
- 21 location pages plus a hub
- 8 industry pages reframed in capability-only language
- Blog, comparison, insight, and use-case routes
- Zero fabricated organization claims, zero pricing site-wide
Outcomes the engagement actually produced.
Roughly 164 URLs go out as static HTML at build time — services, case studies, locations, industries, blog, comparisons, insights, and use-cases.
An 8,800-line globals.css built on Tailwind v4 cascade layers and HDS tokens drives the visual language — no template framework, no UI kit.
Every page ships with llms.txt visibility, JSON-LD schema connections, agent discovery endpoints, and content-signal headers so AI search engines can cite the work cleanly.
Zero off-the-shelf themes, UI kits, or page templates. Every component — including the case-study and location systems — was authored for this site.
Need a marketing site that is also an engineering artifact?
Preisser Solutions builds custom marketing platforms end-to-end — design system, component architecture, AI-native infrastructure. Scoping begins with a conversation.