Sakhonchyk.

Founder. An engine that finds the right person from a plain-language description instead of filters. Shipped as an API and an MCP server — so somebody else's ATS suddenly knows how to search people.

The operating system of a padel club: a site with CRM and intake, a Telegram bot with admins wired in, court-load tracking, a content plan driven by that load, and a corporate offer that tracks who opened it.

Co-founder. An app for online networking and a format for meeting people at events. Built the US market from zero: 20+ deep interviews, 1,000+ other research participants. Closed, pivoted into Pearch.

Technical director of the brand agency. Every marketing fit-out of the stores across the country, working straight with the head office in Panama. Adapted international window designs to local manufacturing and ran simultaneous night shifts in different cities.

Drove traffic and made content for the brand. Four products under it.

Managing partner. A remote team of design, development, 3D and motion — seven people from four countries. Launched the Hyundai Sonata in the UAE. Digitised 70+ Elikor hood models plus a configurator.

Senior account manager. 18 projects at an average budget of RUB 3–5M, five of them with public figures. Shoots with crews of 50+. One project went wide in the media.

From selling industrial ventilation to running construction sites. Two or three sites at once, up to 40 people reporting directly — the most I have ever led.

Time tracking, face recognition, interactive tablets on five sites, electronic data capture. A rollout of about RUB 100K saved RUB 500K+ of payroll in two months.

A product-owner agent that manages me: ranks projects by impact and cost, runs scheduled tasks, swaps my phone wallpaper to match the focus, and reads information back through neuro-interface rules.

A device for children that warns about a meltdown early — early enough for a parent to act. Roadmap: a paper in nine months.

Deeper into this case through the door below.

It parses my screenshot feed and turns findings into rules I then work by. Not a grey prototype — a product I open daily.

An open repository for language learning with a paid support package on top.

An MVP on satellite data. A recent archive. Looking for who actually needs it before building further.

It decomposes a task and hunts the leading solution for each part: GitHub open source → products → and if there are none or they are absurdly priced, how to build it yourself and which services you would need. Recursively, until the ideal.