ST
Sanjeev ThakurSEO Specialist

Case study · This website

This website: built to be found for my own name.

I rebuilt my portfolio as an SEO project in its own right: a fast static site with a connected knowledge graph, dual sitemaps, an llms.txt file and open doors for AI crawlers, so search engines and AI assistants can recognize and cite me.

Built November 2026 Pre-launch · running locallyAstroTailwind CSSTypeScriptLucide iconsSchema.org JSON-LD
9
Indexable pages
2
Client case studies
33
Schema.org types
33
AI crawlers welcomed
Desktop homepage
Desktop screenshot of the Sanjeev Thakur SEO portfolio homepage
Mobile screenshot of the Sanjeev Thakur SEO portfolio homepage
Mobile homepage

Measured, not promised

Mobile Lighthouse scores will be published here after the first run against the live site. Local development builds are not representative, so no numbers until then.

—
Performance
Pending
—
Accessibility
Pending
—
Best practices
Pending
—
SEO
Pending

Structured data on every page was validated with a custom script: valid JSON, one @graph, no dangling references.

More than a typical portfolio

What most personal portfolio sites ship with, compared with what this one does.

FeatureTypical portfolioThis website
Structured dataOne Person snippet, if any33 linked types in one @graph per page
SitemapsPlugin-generated XML onlyXML + HTML sitemap from one page list
AI-readable summaryNoneGenerated /llms.txt
AI crawler policyDefault robots.txt33 AI agents explicitly allowed
Case studiesLogos and vague claimsAnonymized, metric-qualified write-ups
Share image & iconsDefault favicon1200×630 OG image, full icon set, manifest
Themes & motionSingle themeDay / night themes, reduced-motion aware

The challenge

"Sanjeev Thakur" is a common name. When a recruiter or client searches for me, or asks an AI assistant about me, there is no guarantee they land on the right person. A portfolio that only looks good does not solve that. It has to be unambiguous to machines as well as people.

SEO strategy

Treat my own name as the primary keyword and build technical SEO before design: a crawlable static site, one consistent entity for "Sanjeev Thakur, SEO Specialist, Bengaluru", and every discovery file (sitemaps, llms.txt, robots.txt) pointing search engines and AI to the same facts.

Implementation

  1. 01

    Static-first with Astro

    Pages render to plain HTML with almost no client JavaScript, so crawlers and AI fetchers see the full content on the first request.

  2. 02

    One source of truth

    Profile, experience, skills, case studies and the page list live in typed data files. Pages, schema, sitemaps and llms.txt are all generated from them, so they never drift apart.

  3. 03

    Connected knowledge graph

    Every page ships a single JSON-LD @graph with stable @ids: the Person node links to roles, employer, degree, occupation code, case-study articles and Wikipedia-backed topics.

  4. 04

    Per-page metadata

    Unique titles, descriptions, canonical URLs, Open Graph and Twitter cards with a dedicated share image on every route.

  5. 05

    Dual sitemaps

    An XML sitemap for search engines and an HTML sitemap for people, both linked from the footer and robots.txt.

  6. 06

    AI-readable files

    A generated llms.txt summarizing who I am, what I do and how my metrics should be quoted, plus a robots.txt that welcomes AI crawlers by name.

  7. 07

    Clean URL structure

    Short, descriptive, lowercase paths with a breadcrumb trail that matches the BreadcrumbList schema on every inner page.

What's under the hood

Crawl and index

  • Canonical URL on every page
  • XML sitemap with lastmod
  • HTML sitemap for users
  • robots.txt generated from config
  • Semantic, server-rendered HTML

Understanding

  • Single linked JSON-LD @graph
  • Person with roles, credential & O*NET occupation
  • Topics linked to Wikipedia
  • BreadcrumbList on inner pages
  • Article schema for case studies

Experience

  • Responsive from phone to widescreen
  • Day and night themes
  • Respects reduced-motion settings
  • Keyboard focus outlines
  • Icons, never emojis

AI visibility

  • Generated /llms.txt
  • 33 AI crawlers allowed by name
  • Content-Signal: search, AI input, training
  • Disambiguating description in schema
  • Metric qualifiers for accurate quoting
GEO · AI visibility

Do AI assistants know who I am?

People now ask AI assistants, not just Google. So I ask the major AI assistants the same question about me and log what they say, whether they get my identity right, and which sources they cite.

0/6
assistants checked
0/6
found me
0/6
cited my LinkedIn

Baseline in progress: the first round of checks is logged after launch. Every result here comes from a real check, never an estimate.

AI assistant visibility comparison
AssistantMentions meOne clear identityCites my LinkedIn
ChatGPTNot checkedNot checkedNot checked
ClaudeNot checkedNot checkedNot checked
GeminiNot checkedNot checkedNot checked
CopilotNot checkedNot checkedNot checked
Google AI OverviewsNot checkedNot checkedNot checked
PerplexityNot checkedNot checkedNot checked

Where the assistants learned about me

Sources cited, out of 6 assistants checked.

  • LinkedIn0/6

Pre-launch baseline: LinkedIn is the row to watch until this website is live and can be cited.

ChatGPT logo

ChatGPT

Awaiting first check

Not checked yet–/3 signals
Question asked

Who is Sanjeev Thakur, the SEO Specialist from Bengaluru, India?

What it said

Not checked yet. The answer summary will appear here after the first check.

Sources

None logged yet.

  • Mentions meNot checked
  • One clear identityNot checked
  • Cites my LinkedInNot checked
What I take from it

Pending the first check.

What this site already does for AI visibility

  • llms.txt

    A generated plain-text summary of who I am and how my metrics should be quoted.

  • AI crawlers allowed

    robots.txt names 33 AI agents, including GPTBot, ClaudeBot, PerplexityBot and Google-Extended.

  • Entity schema

    One Person node with a stable @id, referenced from every page, linked to roles, degree and topics.

  • Answer-shaped content

    Case studies state results plainly, with exactly what each metric does and does not mean.

  • Consistent identity

    One name, title and location used everywhere on this site and linked to my LinkedIn profile, so assistants can tell me apart.

Re-checked monthly with the same question and logged in src/data/geo-log.ts. Answers vary by session and by whether web search is on, so read this as a trend, not a score. Assistant names and logos are used only to identify each product and belong to their owners. Logos via Lobe Icons (MIT).

Where it ranks for my name

Ranking log starts at launch

Positions for branded queries will be logged here from Google Search Console once the site is live and indexed.

When it was made

  1. 1

    Portfolio foundation

    Nov 2026

    Astro + Tailwind build with About, case studies, contact and day/night themes.

  2. 2

    Homepage redesign

    Nov 2026

    Animated hero, rotating headline, skills ribbon, scroll reveals and a redesigned experience timeline.

  3. 3

    Discovery layer

    Nov 2026

    XML sitemap, HTML sitemap and a generated llms.txt from one shared page list.

  4. 4

    Open to AI crawlers

    Nov 2026

    robots.txt explicitly welcoming 40+ AI and search crawlers, plus content signals.

  5. 5

    Connected knowledge graph

    Nov 2026

    One linked JSON-LD @graph per page: Person, roles, credential, occupation, pages, articles.

  6. 6

    Brand assets

    Nov 2026

    Favicon set, Apple touch icon, web manifest and a 1200×630 share image.

  7. Launch on custom domain

    Next

    Deploy, then submit sitemaps to Google Search Console and Bing Webmaster Tools.

  8. First Lighthouse + AI baseline

    Next

    Record real scores and the first round of AI assistant checks on this page.

  9. First AI citation

    Next

    Track when assistants start naming and citing this site for my name.

Key takeaways

1

Your name is a keyword

Branded search is the one query every recruiter and client runs. Treating my own name as a target keyword shaped every technical decision on this site.

2

Structure beats volume

A small site with a clean, connected entity graph tells search engines and AI more about who I am than dozens of loosely related pages.

3

Measure in public

Publishing a baseline, even an empty one, keeps me honest. Scores and AI checks get added here as real data, never as promises.

Hiring or Need SEO Help?

Want More People to Find Your Website?

Whether you're hiring an in-house SEO Specialist or need help growing organic traffic, I'd like to hear from you.

4+ years in SEO Client work kept confidential Technical & local SEO