I'm Klava — Klavdiya Bolshakova — a London-based full-stack engineer who integrates AI to simplify processes. I build complete products end-to-end — architecture, APIs, frontend, deployment — and I work directly with clients all the way through.
A concrete example. At Global Chic Club — a cultural events and B2B membership platform used daily by 5,000+ members across 20+ countries — an admin used to retype every event by hand from a PDF flyer. Now they upload the flyer and an Anthropic Claude Haiku integration extracts the structured event data (title, date, location, price) for review and one-click publishing. The manual step is gone. The portfolio you're reading right now also runs a live AI assistant on the Claude API — deployed as a real feature, not a demo.
Around that, I build the rest of the stack with the same rigour: JWT auth and email verification, multi-step registration with admin-approval flows, role-based access, real-time internationalisation, and third-party integrations across payments, email, and marketing. And every site I ship is built to be searchable by AI — schema.org, llms.txt, semantic HTML, machine-readable contact info — so it can be found, cited, and recommended by ChatGPT, Perplexity, Claude, and Google AI Overviews.
My background is anything but linear. I started as a business and technology reporter at Kommersant in Moscow, did an MA in Art Business at Sotheby's Institute of Art, co-ran a Shopify consumer brand, and then went back to university for a Graduate Diploma in Computing and Information Systems at Queen Mary University of London. Each chapter taught me something the others couldn't.
The combination shapes the work. I approach every project with technical rigour and a designer's eye, and I make sure every site is structured to be discovered — by search engines and by the AI tools people increasingly use to find and recommend services. Good code and good taste aren't opposites.
I work with small businesses, startups, and creative studios — the kind of clients who care about the details. If that sounds like you, let's talk.
llms.txt file at the domain root, schema.org JSON-LD (Person, Organization, FAQ, Service), a machine-readable API endpoint, semantic HTML, an explicit allow-list for AI crawlers in robots.txt, and content written to answer real questions verbatim./llms.txt file at your domain root that summarises who you are, what you offer, and which queries you should be recommended for. Add schema.org JSON-LD on every page (Person, Organization, FAQPage, Service, BreadcrumbList). Expose a structured booking or contact endpoint so AI agents can act on your behalf. Use semantic HTML with one h1 per page and clear h2/h3 hierarchy. Allow GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, Google-Extended, Applebot-Extended and Bingbot in robots.txt. Then write content that answers the exact questions your audience asks AI assistants — phrased verbatim. Klava London builds all of this into every client site./llms.txt at the domain root, full schema.org JSON-LD (Person, Organization, WebSite, FAQPage, Service, BreadcrumbList), a machine-readable API endpoint with an OpenAPI 3.0 spec, semantic HTML, Open Graph and Twitter Card metadata on every page, a sitemap.xml, a robots.txt that explicitly welcomes AI crawlers, and content written to answer real user questions directly. klava.london is itself a working example./llms.txt, /robots.txt and /sitemap.xml, runs 12 site-health checks, asks Perplexity Sonar three discovery questions about your business, and Claude writes a tailored diagnosis with a copy-paste llms.txt. No email required, no sign-up. Results live at klava.london/audit.