Cases

Five pieces of work that show how I think. Older ones are the most verifiable; newer ones are the most relevant.

Running a learning marketplace, 2023 → today

Online learning marketplace · CEO · growth, economics, team

I took over as CEO in 2023, at a crisis point for an already-mature product: a course marketplace with 1.5 million monthly users and more than 20,000 courses, and no outside investors. In that time the number of authors earning income on the platform has nearly tripled, and the income they earn has more than tripled. The product now serves noticeably more learner segments than it did — we grow wide, not just deep. The business is self-sustaining.

The team has grown in capability three ways at once: hiring and renewal, a working model with external contractors, and targeted use of AI tools where they actually strengthen a person’s output rather than replace it. In 2025 we revived the company’s offline conference for education practitioners after a six-year pause; it now brings brand effect, new authors, and new courses created in partnership with authors who were already on the platform.

Cohort-based bootcamps from a self-paced product

Hyperskill · 2023 · new business line

Self-paced learning had a ceiling on both outcomes and price. Designed the cohort format, built the tooling, structured the curriculum, hired the team and made the first sales. The line still runs as sequential cohorts, with costs that grow much more slowly than enrolment.

Example: go.hyperskill.org/ai-engineer

A social network for kids, rebuilt to 30K MAU

Kodland · 2021–2022 · rebuild, community, team formation

Inherited a rebuild with one ticket and three weeks. Built the roadmap, the team, the moderation framework and a contest calendar. Six months later: 30,000 monthly active users, 3.5 pages per session, engagement up 50%. When retention dropped after a change to the course system, ran over a hundred interviews with children and parents and turned them into a new product concept.

Proof: hub.kodland.org

Recommendations that moved the top line

MTS Big Data · 2019–2021 · ML product, growth

Personalised offers on the homepage: +80% conversion for authenticated users. Device recommendations across email, SMS, push and banners: +1% to the retail customer base. Email segmentation: +40% unique opens.

Schema.org VideoGame — shipping a web standard

Yandex · 2013–2015 · standards, cross-company work

Yandex needed structured data for game descriptions and nothing suitable existed in Schema.org. I proposed the vocabulary, took it through the international working group (Google, Microsoft, Yahoo, the W3C community), handled the “what even is a game” definition debate, and shipped it into Yandex’s production webmaster tools. The vocabulary is now used on over a thousand sites, and Yandex built rich search results on it.

Proof: schema.org/docs/about.html · github.com/schemaorg/schemaorg/issues/169