AI Engineer, full-stack: LLM apps and automation

Work that shows how I build

• Growth report: a structured LLM pipeline

  A free tool on this site: enter a website and get a report built from the fetched pages, live speed measurements and AI-written sections.

  • JSON-schema output: strict structured outputs on OpenAI, or the schema in the prompt plus local validation and one corrective retry on other OpenAI-compatible models.
  • A quote from the analysed site is kept only if it appears word for word on the fetched pages.
  • Every fetch, including each redirect, is refused if it resolves to a private or internal address.

  TypeScript · Next.js · OpenAI-compatible APIs · Supabase

  See it livePrivate repository

• This website

  A Next.js App Router site in English, French and Arabic, with Arabic rendered right to left.

  • Middleware maps public URLs to locales and keeps English at the bare path; typed dictionaries make a missing translation fail the type check.
  • A strict Content-Security-Policy and HSTS on every route.
  • Canonical, hreflang and JSON-LD on every page, and a sitemap that lists every locale.

  Next.js 16 · React 19 · TypeScript · Tailwind CSS

  See it livePrivate repository

• IdeaSignal

  Turns a startup idea into problem-language Reddit searches, collects the threads and has an LLM score the demand evidence.

  • Scores pain, workarounds, switching and spending signals, and penalises single viral threads and self-promotion.
  • Runs on any Anthropic-compatible endpoint, with a mock-data mode so the interface runs without keys.

  Next.js · TypeScript · Reddit API · Anthropic SDK

  Code (opens in a new tab)