Growth Marketer building AI workflows for real campaign work — not demos, not templates. Everything here has been tested in production at a top-10 fintech app in Southeast Asia.
I share what I actually use: prompts that cut campaign brief time from 3 hours to 20 minutes, cohort forecasting that flags where retention breaks before it costs budget, MEU planning that works backward from targets instead of forward from guesses.
Prompts and Python agents for growth marketing work. Campaign brief, MEU planning, A/B test analysis, multi-agent workflows. Installable as Agent Skills via npx skills add thaolst/ai-growth-agents-for-marketers.
MCP server encoding growth marketing domain knowledge as callable tools. Campaign design, retention analysis, churn prediction, channel mix, cohort forecasting. pip install growth-mcp. 9 tools, 71 tests.
Prompt library from live campaigns. Voucher design, user segmentation, game mechanics, retention loops. 40+ prompts tested in production.
Personal AI agent on Telegram. Flight search and price comparison in natural language. Claude Sonnet + SerpAPI. 55 stars · 67 forks.
| Workflow | Before | After |
|---|---|---|
| Campaign brief | 3 hours | 20 min |
| A/B test analysis | 3 hours | 15 min |
| MEU planning | 2-3 days | 30 min |
| Campaign history query | 30 min | 30 sec |