Choose Your Starting Point
Each part builds on the previous one. Start from Part 1 if you're new, or jump to any section if you're already familiar with the basics.
01
The Big Picture
Define useful AI assistance, separate facts from inferences and proposals, and map the trust boundaries before connecting a production account.
02
What You Can Achieve
8 real-world use cases with interactive animated demos. From ad testing at scale to competitive intelligence to broken stock recovery.
03
Generative vs Agentic AI
Compare conversational drafting with tool-using workflows, then design explicit data, authorization, approval, and execution boundaries.
04
Foundation & Setup
Build a synthetic sandbox with separate identities, managed secrets, allowlisted tools, schemas, dry runs, audit events, and tested stops.
05
Skills & SOPs
Turn reviewed SOPs into versioned Skills with narrow scope, declared inputs, evidence checks, failure states, evaluations, and owners.
06
Six Controlled Workflow Patterns
Apply the architecture to creative, launch proposals, budget recommendations, monitoring, reporting, and scaling scenarios without granting the model authority.
07
Operations & Beyond
Daily AI routines, chaining agents, building custom skills, MarTech as a career, the data opportunity, and important disclaimers.
💡 New Here?
Start with Part 1: The Big Picture for the evidence model and safety boundary used throughout. You can learn the concepts without platform access; implementation requires qualified reviewers and an isolated test environment.
Built by @itsmazinzaki — AVAMARTECH
Reviewed educational edition — 31 August 2026