Summary: SEO basics for technical AI/ML content: canonical URL hierarchy prevents duplicate content, structured data helps rich snippets, internal linking builds topical authority. For PhD portfolio, expertise demonstration > traffic volume.

Canonical URL Strategy (Core)

Owned Domain (ahsan.ai/blog/...)
       ↓ rel="canonical"
Substack (imported, canonical set to owned)
       ↓ noindex or canonical
LinkedIn Article (condensed, link to canonical)
       ↓ noindex
X/Twitter Thread (no link in tweets, link in reply)

Rule: One canonical URL per piece of content. All others point to it.

Technical SEO Checklist

Element Implementation
Canonical tag <link rel="canonical" href="https://ahsan.ai/blog/slug" /> on all versions
Meta description 150–160 chars, includes primary keyword, compelling
Title tag <60 chars, keyword front-loaded, unique
Heading structure H1 → H2 → H3 (single H1 per page)
Structured data BlogPosting schema (JSON-LD)
Internal links Wiki-style <span class="wikilink unresolved">concept-slug</span> → rendered as links
Image alt text Descriptive, includes keyword naturally
URL slug kebab-case, descriptive, <60 chars

Structured Data (BlogPosting Schema)

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "Why Flash Attention Matters for LLM Training",
  "description": "TL;DR summary...",
  "image": "https://ahsan.ai/images/flash-attention-diagram.png",
  "author": {
    "@type": "Person",
    "name": "Ahsan Umar",
    "url": "https://ahsan.ai"
  },
  "publisher": {
    "@type": "Organization",
    "name": "Ahsan AI",
    "logo": { "@type": "ImageObject", "url": "https://ahsan.ai/logo.png" }
  },
  "datePublished": "2026-07-16T10:30:00Z",
  "dateModified": "2026-07-16T10:30:00Z",
  "mainEntityOfPage": "https://ahsan.ai/blog/why-flash-attention-matters"
}

Keyword Strategy for AI/ML Researcher

Keyword Type Examples Strategy
Primary (high intent) "FlashAttention implementation", "LoRA fine-tuning tutorial" Target in H1, first paragraph, slug
Long-tail (specific) "Flash Attention 2 vs 1 benchmark RTX 4090" Target in H2/H3, deep sections
Conceptual "attention mechanism explained", "what is LoRA" Build topical authority via concept pages
Paper-based "DeepSeek V4 CSA HCA attention", "HRM paper explained" Paper breakdown pages

For PhD portfolio: Rank for "your name + research area" and "your papers" > generic keywords.

Internal Linking (Topical Authority)

Concept pages (wiki) ↔ Blog posts ↔ Paper breakdowns
       ↓                      ↓
Projects ↔ Research ideas ↔ Resume

Every blog post links to:

  • 3–5 concept pages (<span class="wikilink unresolved">concept-slug</span>)
  • 1–2 paper breakdowns
  • 1 project page (if applicable)
  • Resume (auto-generated)

Content Quality Signals (E-E-A-T)

Signal Implementation
Experience "In my implementation..." + code repo links
Expertise Equations, architecture diagrams, benchmarks
Authoritativeness Citations to papers, professor mentions, lab affiliations
Trustworthiness Honest limitations section, reproducibility notes

Substack Import SEO

Setting Value
Canonical URL Set to owned domain URL
Meta description Auto from first 160 chars (override if needed)
Social preview og:image = blog hero image (1200×630)
Indexing Allow (canonical handles duplicate)

Related Concepts

Sources