STACK

YURI RYBAK · SERVICES, INFRA, DATA, PROJECTS
The toolkit and infrastructure I build websites and apps on. AI-generated UI, GitHub for source, Vercel for hosting, Hetzner for heavier data work, and a rotating set of AI / media / social APIs depending on the project.
DEFAULT WEBSITE BUILD ORDER
STEPSERVICEUSE FOR
1. UI scaffoldv0Generate Next.js / React UI from prompts
2. RepoGitHubSource control; also small JSON state via Contents API
3. HostVercelDeploy + serverless functions + cron
4. DB / authSupabasePostgres + auth + storage when GitHub-as-DB isn't enough
5. AI genAnthropic / OpenAI / Fal / Replicate / RunwayText, image, video, audio
6. Email + cronResend + Vercel cronTransactional email + scheduled jobs
7. QAPuppeteerVisual QA, login-gated scrape, e2e flows
FAL VS REPLICATE
HOUSE RULES
RULEWHY
No hardcoded credentials in sourceEnv vars only, mirrored into Vercel per project
No Vercel Blobs for small JSON stateCDN cache makes admin writes appear stale; GitHub Contents API instead
Date-stamped project folders (NAME-YYYY-MM-DD)Convention; keeps the timeline searchable
Verify the deployed artifact contains the changeNever assume cp / build worked
The services I reach for first when building. Grouped by what they do. Pricing tier and exact integration detail vary per project.
Two Hetzner Cloud servers in Helsinki: a production API box that powers the Everysong music data API, and a separate ingest / staging box for building the indexes the API serves.
Everysong is the music data backbone for Driftwave, Automash, and other music projects. SQLite via better-sqlite3, served from the production API box.
All yuribeats GitHub projects. Each card links to the repo; live URLs link to the deployed site where there is one. Public and private repos both shown — private ones are visible by name only.
All Vercel deployments. Each card links to the live site; repo column maps back to the GitHub source.