Roman Makuev

Experience

For the last decade I've worked at the intersection of search and engineering — starting in classic on-page SEO and moving into the mechanics of how modern retrieval systems actually read, rank, and cite content. That shift is where most of my work now lives: vector distance between pages, passage-level information gain, entity coverage, and the retrieval constraints that decide whether a page shows up in an AI answer or gets skipped.

The work is measured, not asserted. I've tracked 180 domains through a single core update to isolate what actually moved rankings, and analyzed information gain across roughly 4,000 top-ranking URLs to see what genuinely new information separates cited pages from the rest.

Expertise & Proof

At Neon I architect the systems behind that work: crawlable, render-safe front-ends; search demand mapped into information architecture before a line of code is written; and the analytics platform — Algorithm — that lets clients see what search engines see in real time, instead of relying on delayed third-party scrapes.

Algorithm is where the theory becomes a tool. It runs live vector and retrieval analytics — query mapping, semantic coverage, information-gain scoring — the same intelligence layer behind Neon's client engagements. Everything I publish comes out of that practice, not the other way around.