Anton K.

I build rule-based automation for Meta (Facebook) ad accounts using the Marketing API and Google Apps Script. Typical setup: auto-pause of burning creatives by CTR/CPA/spend thresholds, budget scaling by ROAS, daily reporting to Google Sheets and Telegram, and a human-approval gate before anything executes on live spend. Shopify integration: orders, revenue and costs pulled via the Shopify Admin API into the same Sheet as Meta spend, so ROAS and profit come from real store data, not just pixel-reported conversions. Deployed via a dedicated app and System User inside the client's own Business Manager — no access sharing, no App Review bottleneck. Currently running this stack in production for a multi-market advertiser (3 markets, live budget). I also build LLM-based tooling with the same principle — AI proposes, a human approves: e.g. an agent that turns a client-feedback Telegram channel into prioritized tasks and drafts code fixes for review. I work in a narrow niche on purpose: Meta Marketing API, Shopify store data, Apps Script, ad-account automation, reporting, and AI agents around these workflows. If your task is outside that, I'll say so.