I help brands stay visible when the answer is no longer a list of blue links — but a paragraph generated by ChatGPT, Perplexity, Gemini or Google AI Overviews. AI SEO, GEO, AEO and LLMO, engineered as one strategy: machine-readable trust, entity clarity, citation worthiness.
— (01) The role
An AI Search Consultant closes the gap between how your brand exists today and how AI systems need to read you tomorrow.
Classical SEO optimised pages for ranking on a results page. AI search optimises something deeper: the entity behind your brand, the structure of your content, and the signals of trust that decide whether a large language model is willing to cite you.
The work brings together SEO, GEO (Generative Engine Optimization), AEO (Answer Engine Optimization), LLMO (Large Language Model Optimization), entity SEO, semantic SEO, schema markup, structured data, citation engineering and retrievability — into a single, coherent visibility programme. Not ten dashboards. One strategy.
If you have a website that ranks but isn't quoted, a brand that competitors mention but AI assistants don't, or content that's read by humans but invisible to retrievers — this is the discipline that fixes it.
— (02) Who this is for
Engagements are tailored, not templated. The fit is consistent though — teams that already understand SEO and now need an AI search layer.
SaaS, fintech, devtools, professional services — companies whose buyers research with AI before vendors.
Early-stage and growth founders who want category visibility without burning years.
Specialists who need their name, not just their site, to be cited inside AI answers.
DTC and retailers fighting for product-question intent inside AI Overviews.
Editorial teams whose content is being trained on but not credited.
Multilingual brands needing entity coherence across languages and markets.
— (03) Core programs
Each program can stand alone or compose into a full AI search strategy. Most clients start with the first two and expand from there.
The umbrella program.
A 90-day visibility plan that maps your current AI footprint across ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews, then prioritises the structural, content and citation work that will move you from invisible to recommended.
Make your brand a first-class entity.
Disambiguate your brand, founder and offerings as entities the open web — and downstream LLMs — can resolve unambiguously. Wikipedia/Wikidata signal hygiene, sameAs networks, knowledge panel readiness, multilingual entity coherence.
Generative, answer and language-model optimisation.
The trio that defines AI-era search. GEO engineers citation worthiness inside generative answers. AEO captures zero-click question intent. LLMO works the broader layer of how language models perceive, attribute and recommend you.
Content that retrieves cleanly.
Information architecture, page templates, internal linking and answer-block patterns designed for retrieval. Content that earns AI citations isn't just well-written — it is well-structured for a model that scans, not reads.
A type system for the open web.
JSON-LD schema, FAQ/HowTo/QAPage markup, Article and Person graphs, and the citation hygiene (sources, dates, authority links) that signal trust to both Google and AI systems sourcing answers.
Skill-up your in-house people.
Workshops and cohort training for SEO, content and product teams transitioning from traditional SERP optimisation to AI-first discoverability. Practical playbooks, prompt-driven testing, measurable assignments.
— (04) Problems solved
If two or three of these are true, the consulting work pays for itself fast.
Invisible in AI answers.Your competitors are quoted in ChatGPT and Perplexity. Your brand isn't.
Weak entity recognition.Search engines can't reliably tell who you are, what you do, or where one ends and another begins.
Content not built for retrieval.Long-form posts that read well, but are skipped by models scanning for answer blocks.
Poor AI Overview performance.You appear in classic SERPs but never inside the AI Overview that now sits above them.
Unclear service signals.Your offering is hard for a machine — and increasingly, a buyer — to summarise in one sentence.
No machine-readable trust.No schema, no source attribution, no entity graph. Nothing for an LLM to anchor on.
— (05) Why work with me
Most agencies are still optimising for the SERP of 2018. The work I do is purpose-built for the answer engines of now.
I'm an independent AI Search Consultant based in Belgrade, working with brands across EMEA, DACH, the Western Balkans and the global remote market. The practice has been independent since 2013 and intentionally small — fewer than eight clients per year, no juniors, no handoffs.
What clients hire me for: a fluent grasp of classical SEO and what is actually happening inside large language models. Entity reasoning. Schema as a type system. Citation engineering as a craft. Multilingual visibility — done with the same care in Serbian as in English.
What I refuse to deliver: dashboards full of vanity metrics, AI-buzzword strategy decks, content factories. The output is always tied to a measurable visibility shift.
— (06) Framework
Every engagement, regardless of scope, runs through the same disciplined sequence. No black box, no surprises.
Benchmark current visibility across Google, AI Overviews and the four major LLMs. Map citations, mentions, entities, retrievability and competitive moats.
Define the entity graph: brand, founder, services, products, sub-topics. Resolve ambiguities, align sameAs networks, plan multilingual coherence.
Rework information architecture and page templates so the highest-value answers retrieve cleanly. Answer blocks, FAQ structure, internal linking.
JSON-LD schema (Person, Organization, ProfessionalService, Article, FAQPage), source attribution, freshness signals, citation outreach where it earns lift.
Your team or mine. Playbooks, briefs, technical specs, prompt-based test suites. We ship in weekly increments, not quarterly waterfalls.
Custom dashboards tracking AI Overview share, LLM citation share, branded prompt accuracy and downstream pipeline. Iterate weekly. Compound monthly.
— (07) Outcomes
Outcomes, not outputs. Here's what shifts when the work is done well — measured against pre-engagement baselines, not vanity metrics.
Branded prompts return correct, complete and quotable summaries across ChatGPT, Perplexity, Gemini and Claude.
Your brand resolves to a single, unambiguous entity across Wikipedia, Wikidata, Google Knowledge Graph and downstream LLMs.
Your content earns its place as a source — quoted with attribution, not paraphrased anonymously.
Schema, freshness, sourcing and authority links render your domain as a high-confidence source to any retrieval system.
One sentence a model — or a buyer — can repeat back. No ambiguity, no overlap, no slop.
The structural work compounds: every new piece of content, every new market, every new product slots into a system already engineered to be found.
— (08) FAQ
Eight short answers to the questions clients tend to ask in a first call.
— Ready when you are
If your brand belongs in AI-generated answers — and isn't there yet — let's talk. The practice is currently booking Q3 / Q4 2026 engagements.