Free Guide: The Self-Learning Agent Setup (memory + feedback loop, 23 pages)

Your AI agent forgets every correction you give it. This guide fixes that.

Anyone can pay for the same model you use. What nobody else can buy is (1) what your agent knows about your business and (2) the record of every correction you've given it. This free 23-page field guide shows you how to set up both, the way I run my own one-person business in Korea: agents draft, I approve or fix from my phone, every decision goes into a database, and the next draft reads that history first.

Inside:

• The knowledge layer: brand files vs. live database vs. searchable content (RAG), and why prices should never live in a text file
• The learning loop: store, generate candidate rules, score, compare, promote or drop
• Rules vs. memory, and making past feedback searchable by meaning
• The Telegram approve / reject / edit setup, and how to check it's actually saving
• A hypothetical week at a flower shop, step by step
• A starter database schema (Postgres / Supabase), 11 copy-paste prompts, and a 7-day plan

No code required. You need to know what to ask for, in what order, and what to check.

Free download. It's two of the eleven steps from Just Say "Do It": The AI Agent Playbook.

The whole system on video: my course has 68 lessons (13h 26m), recorded in Korean with an English translation: ssapable.com/courses/ai-agent