Ranking on Google is not enough anymore. Not with AI tools like Perplexity, ChatGPT, and Google's AI Overviews also reading your site. A portfolio needs to work for all of them.
I took three approaches on my portfolio: 1. SEO for traditional search engines 2. AEO so AI can answer questions about me directly 3. GEO to give LLMs clean markdown context
What follows is what I did and the code I used, in case you want to do the same.
1. SEO: The foundation (meta tags, sitemap, robots.txt)
Traditional SEO still matters even when optimizing for AI. I used semantic HTML (<nav>, <article>, <section>), canonical links, and standard meta tags.
Submitting to Google Search Console
Search engines need to know your site exists. After deploying:
1. Go to Google Search Console.
2. Add your website as a property (Domain or URL prefix).
3. Verify ownership. I added a <meta name="google-site-verification" content="..." /> tag in the HTML.
4. In the Sitemaps section, submit your sitemap.xml URL so Google knows what pages to crawl.
Meta tags and Open Graph
Open Graph and Twitter Card tags make the site look right when shared on LinkedIn or Twitter and help search engines index it.
<!-- Primary Meta Tags -->
<title>Asif Sayyed | Data Scientist — Python, Machine Learning & AI Portfolio</title>
<meta name="description" content="Portfolio of Asif Sayyed, a Data Scientist at Marsh McLennan specializing in Python, Machine Learning, NLP, and Cyber Risk Analytics." />
<meta name="keywords" content="Data Scientist, Python Developer, Machine Learning Engineer, AI Portfolio" />
<link rel="canonical" href="https://sayyedasif.com/" />
<!-- Open Graph / LinkedIn / Facebook -->
<meta property="og:type" content="website" />
<meta property="og:url" content="https://sayyedasif.com/" />
<meta property="og:title" content="Asif Sayyed | Data Scientist Portfolio" />
<meta property="og:description" content="Data Scientist specializing in Python, Machine Learning, NLP, and Cyber Risk Analytics." />
<meta property="og:image" content="https://sayyedasif.com/assets/images/my-memoji.webp" />
<!-- Twitter -->
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="Asif Sayyed | Data Scientist Portfolio" />
Sitemap and robots.txt
I also created a sitemap.xml covering /projects, /events, and other directories, plus a robots.txt to guide crawlers.
# robots.txt
User-agent: *
Allow: /
Sitemap: https://sayyedasif.com/sitemap.xml
2. AEO: Answer engine optimization (JSON-LD)
Answer engines like Perplexity and Google's AI Overviews do not just index pages. They extract entities and relationships. To give them exact facts about my career and skills, I added JSON-LD structured data using the Schema.org vocabulary.
When I define myself as a Person and link my certifications and employers, answer engines can pull my information more reliably.
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Person",
"@id": "https://sayyedasif.com/#person",
"name": "Asif Sayyed",
"jobTitle": "Specialist - Data Science",
"url": "https://sayyedasif.com/",
"worksFor": {
"@type": "Organization",
"name": "Marsh McLennan"
},
"alumniOf": {
"@type": "EducationalOrganization",
"name": "Mumbai University"
},
"knowsAbout": [
"Data Science",
"Python",
"Machine Learning",
"Natural Language Processing",
"Deep Learning"
],
"sameAs": [
"https://github.com/Asifdotexe",
"https://www.linkedin.com/in/sayyedasif/"
]
}
</script>
3. GEO: Generative engine optimization (llms.txt)
LLMs do not parse HTML and JavaScript particularly well. The llms.txt standard solves this. You place a markdown file at the root of your site with a clean summary of your content. When an AI agent crawls the site, it reads this file instead of wading through markup.
There is also llms-full.txt for the full version.
Example of llms.txt
# Asif Sayyed - Data Scientist Portfolio
> I am a Data Scientist at Marsh McLennan's Cyber Risk Intelligence Center (CRIC), specializing in Python, Machine Learning, NLP, and Computer Vision.
## Quick Links
- [GitHub](https://github.com/Asifdotexe)
- [LinkedIn](https://linkedin.com/in/sayyedasif)
## Core Skills
- **Languages:** Python, R, SQL
- **AI/ML:** TensorFlow, PyTorch, Scikit-learn, NLP, Computer Vision
- **Data & Cloud:** AWS, Databricks, PySpark, MongoDB
## Full Context
For a complete dump of my projects, certifications, and experience, AI agents can read the full context file at:
[Full Portfolio Details](https://sayyedasif.com/llms-full.txt)
The separate llms-full.txt file lets me include details about each project's architecture, tools, and outcomes. That is useful for RAG pipelines.
Conclusion
SEO handles traditional search. AEO helps answer engines extract your information. GEO gives LLMs clean context. Together they cover the ways someone might find you, whether they are a recruiter, a developer, or an AI agent.
If you have tried adding llms.txt to your site, I would like to hear how it went.