avaluev.github.io

Personal site of Alex Valuev — Senior AI Product Manager and career coach. Live at https://avaluev.github.io/.

This repo is the source of the deployed site. Static HTML, no build step at runtime, no JavaScript framework, no tracking scripts. The only generators are two Python scripts that produce SEO assets and the OG preview image.

Quick links

Live site	https://avaluev.github.io/
LinkedIn	https://www.linkedin.com/in/valuev/
Email	mailto:valuev.alexandr@gmail.com
Telegram (direct)	https://t.me/asnkt
VALUEV CAREER channel	https://t.me/itcareertech
YouTube	https://youtube.com/@itcareertech
Flagship public research	https://avaluev.github.io/padel-market-analysis/

What ships in this repo

.
├── index.html           # Landing — hero, signature stats, featured work, CTAs
├── about.html           # Long-form bio, career history, leadership philosophy
├── projects.html        # Public projects grid
├── coaching.html        # VALUEV CAREER service page
├── contact.html         # Direct contact methods
├── 404.html             # Friendly not-found page
├── favicon.svg          # Brand mark
├── topnav.css           # Shared navigation styles
├── manifest.webmanifest # PWA manifest (generated)
├── robots.txt           # AI-crawler-aware allow/disallow (generated)
├── sitemap.xml          # XML sitemap (generated)
├── llms.txt             # llmstxt.org index (generated)
├── llms-full.txt        # Full-text retrieval corpus (generated)
├── feed.xml             # RSS feed (generated)
├── humans.txt           # Human credit file (generated)
├── og-default.png       # Social-card image (generated)
├── .well-known/
│   └── security.txt     # RFC 9116 security policy (generated)
└── scripts/
    ├── build_seo_assets.py   # Idempotent SEO-asset generator
    ├── build_og_image.py     # Pure-stdlib PNG generator for the OG card
    └── check_quality.py      # 12-check unified content + SEO quality gate

Quality bar

Every page must pass the unified gate before merge. The gate runs in CI on every push and pull request, and locally via make audit.

Per-page checks:

• exactly one <h1> per page
• 40-60 word <p class="summary"> citable summary lead under H1
• full meta package: title, description, canonical, og:title, og:description, og:url, viewport, robots
• valid JSON-LD with @context: "https://schema.org"
• consistent five-link top navigation across all pages
• no broken internal links
• no marketing badges, run-IDs, or jargon (kill experiment, north star, red team, etc.)
• every <img> has alt; warning if missing width (Core Web Vitals)
• no leaked Markdown auto-link <URL> syntax

Site-wide checks:

• every well-known SEO file present and non-empty (robots.txt, sitemap.xml, llms.txt, llms-full.txt, feed.xml, humans.txt, manifest.webmanifest, favicon.svg, .well-known/security.txt)
• sitemap.xml is parseable
• llms.txt starts with an H1

Run all checks:

make audit            # build assets + run quality gates
python3 scripts/check_quality.py --list   # list available checks
python3 scripts/check_quality.py --json   # machine-readable output

Local development

make install          # install ruff / mypy / pytest
make build            # rebuild og-default.png + SEO assets
make audit            # full pre-merge audit
make serve            # serve on http://localhost:8000

The build step is idempotent. CI verifies that make build followed by git diff produces no changes — this catches drift between the source and the generated assets.

Toolchain provenance

The quality-gate scripts and the AI-search-optimisation rules were lifted from the avaluev/padel-market-analysis repo, which is the reference implementation for this engineering bar across all of Alex's public properties.

License

MIT. Reuse, adapt, and remix freely.