How Topicary keeps structure without making teams write XML
Vlad Kuzin, founder of Topicary, built an AI-native documentation platform for small teams that need reusable content without an XML-first workflow. The conversation explores how practical comparison content brings in prospective customers, why Kuzin dropped the XML requirement, and why AI still needs human oversight to keep documentation accurate.
Start by turning the piece of content that appears in the most places into a reusable block. Treat AI as a drafting aid, not as the owner of documentation accuracy.
What’s the short story of how your business started, and what makes it genuinely different from the other options your customers have?
I spent about 15 years in technical writing and UX content, and I kept seeing the same problem. Small documentation teams are stuck between two bad options. A wiki is easy, but the same warning gets copied into dozens of pages and drifts apart. An enterprise content system fixes that, but it costs a lot and expects you to work in XML or DITA. I built Topicary for the gap in between. You write in a normal block editor and reuse content by block, and the system shows every page that uses a block before you change it. Two things set it apart. AI output is checked against your content model before it goes in, and subject experts can review through a link with no login.
How did you win your first customers, and where does most of your new business come from today?
The channel I’ve put the most effort into is writing. I publish plain comparisons and pricing explainers on topicary.com, and posts on DEV and LinkedIn about content reuse and AI in documentation. I cite each competitor’s own pages and date my checks, so people can verify what I say. The people who find those pages are already looking for a way out of a wiki or an XML tool, which is why I put the work there.
What’s one decision that changed the direction of your business, and one mistake that taught you the most?
The decision that changed Topicary most was dropping the XML requirement. Structured authoring has real benefits, like reuse, conditions and variables, but most small teams never get them because of the markup. When I looked at 59 practitioner voices, 47% said XML was unnecessary and another 31% disliked it but felt stuck. So Topicary keeps the structure and drops the markup, and still exports to DITA, Markdown, Word and XLIFF. The mistake that taught me the most, from years in this field, was treating documentation as something you write once. Docs go wrong as the product changes. Now, when a pull request merges, Topicary drafts a doc update linked to it, so the change is reviewed while it’s fresh.
What has changed in your industry or in what your customers expect over the last couple of years, and how have you adapted?
AI changed two things. People now read documentation through AI assistants as well as on pages, so it has to be accurate and structured enough for machines to use. And AI drafts quickly but breaks at scale. If a security warning appears in 40 articles and AI updates one, the other 39 are now wrong. I adapted by making reuse the base, with one block feeding many pages and tracking of where each one is used. AI output is validated against the content model before it’s inserted, and Topicary has an MCP server so outside AI tools can query how documentation components relate.
If you were advising someone starting out in your industry today, what would you tell them to do first, and what would you tell them to ignore?
First, pick the one piece of content that appears in the most places, like a warning, a legal notice or a setup step, and write it once as a reusable block. You’ll feel the benefit within a week. Then ignore two things. Ignore the idea that you need a full enterprise system and XML on day one. And ignore tools that promise AI will keep your docs current on its own. At scale it updates one page and misses the rest, so someone still has to own accuracy.
Thank you to Vlad Kuzin and the team at Topicary for sharing what actually worked with Leaders Perception readers.
Want to tell the story behind your business?
Leaders Perception publishes interviews with business owners and leaders every week. Five questions, about ten minutes, no calls. It is free and every completed interview gets published.