Nano AI - My Bilingual AI Implementation Company Website
Bilingual English/Arabic website for Nano AI, the Gulf-focused AI implementation company I founded, with 16 service pages, pricing, software pages, blog, glossary and structured SEO data.
Client
Date
November 2026
Duration
4 months
Technologies
11 tools
Technologies Used
Next.js
React
TypeScript
Tailwind CSS
Prisma
PostgreSQL
next-intl
Zod
Framer Motion
Vitest
Vercel
Project Overview
Nano AI is the Arabic-first AI implementation company I founded. It serves companies in Saudi Arabia, the UAE, Egypt and the wider Gulf. I built its web presence from an existing Next.js starter into a content-driven marketing site in English and Arabic with full right-to-left support. The site presents 16 fixed-scope services such as WhatsApp AI agents, Arabic voice agents, AI automation and consulting, along with a pricing page, industry and location landing pages, two ready-made software products, a bilingual blog, a glossary, comparison pages and free tools. Pages are generated from typed content files, so the team adds or edits a page without touching layout code.
Challenge
The company sells many services to several markets in two languages, and its growth plan depends on search and AI answer engines finding it. That meant a large number of pages that all needed correct Arabic RTL layout, hreflang pairs, canonical URLs, structured data and consistent pricing. The starting point was a general dashboard boilerplate with authentication, sockets and heavy libraries that a marketing site should never load. Content also had to stay honest: no invented client names or numbers, and prices that match across pages and schema.
Solution
I stripped the dashboard and auth layers and kept a content-only Next.js App Router site with a locale route group. Content lives in typed TypeScript files validated by Zod schemas, with one template per page type: service, industry, location, software, glossary, comparison, tool and blog post. A single SEO config produces metadata, hreflang, sitemaps, llms.txt and JSON-LD for Organization, Service, FAQ, Breadcrumb and Person. Contact requests are saved as leads in Postgres through Prisma with a rate limit and honeypot, and every call to action routes to WhatsApp or a booked scoping call. A shared design system gives the site a dark theme in the brand colours with a light option.
Results & Impact
The site is live in English and Arabic with 16 service pages, 13 industry pages, 13 location pages, 60 glossary entries, 11 comparison pages, 18 blog posts, three interactive tools (ROI calculator, WhatsApp AI cost calculator, accessibility scanner) and two software product pages. Pricing is fixed-scope and shown in USD on one page that matches the schema data. A Bing site auth file, an IndexNow script and an llms.txt route are in place for indexing. Editing a page means changing one content file, and the build checks it against the schema before deploy.
Key Features
Project Gallery
