What I actually do
- Research comes first. Manuals, vendor docs, forums and open knowledge usually contain most of the answer. I look there first, then check what I found.
- The test is whether it still works later. I measure the work by whether the team uses the thing a month later, and whether someone else can maintain it.
- I join methods that were not designed together. Most of my work connects an automation engine, a learning platform, a monitoring stack and an AI assistant. The value is in how they fit.
I’m finishing a Bachelor of Commerce at the University of Victoria (Gustavson School of Business). I have worked in business analysis, AI enablement and technology roles for a regional advisory firm and a national innovation organisation. I speak English, Ukrainian and Russian. I’m based in British Columbia and open to Vancouver, Calgary or remote work.
The CV has the detail. How I work explains the process.
Who helped build the site
I designed, wrote and built the site, working with AI coding assistants (Claude, ChatGPT/Codex, DeepSeek) under my direction and review. These open-source projects and tools did much of the work:
| Project | Used for | Licence |
|---|---|---|
| Hugo | Static site generator | Apache 2.0 |
| three.js | The network field on the home page | MIT |
| Tone.js | Optional interface sounds | MIT |
| h5p-standalone and H5P | Playing the course activities | MIT |
| IBM Plex | Typefaces | OFL 1.1 |
| draw.io | Every diagram, self-hosted | Apache 2.0 |
| Grafana and Prometheus | The monitoring panels on the homelab page | AGPL 3.0 / Apache 2.0 |
| n8n | The contact-form automation | Sustainable Use License |
| Crawl4AI | Collecting public sources for the research case | Apache 2.0 |
| ONLYOFFICE Docs | Rendering the PDF CV | AGPL 3.0 |
| python-docx | Building the DOCX CV | MIT |
| Earshot | Giving the AI ears: listening to and measuring the interface sounds and music | MIT |
The site is static: no cookies, no third-party trackers, no fonts loaded from other servers.