What’s in Private AI Certified — Expert

Knowledge Track

What’s in Private AI Certified — Expert

The application tier. Thirteen papers that take the concepts from Basic and turn them into judgment — real customer scenarios, real economics, real proof points.

27

Papers

Basic

Prerequisite

~6 h

Badge validity

1

When AI Stopped Being an Engineering Project

The business-conversation reframe.

2

How the Data Gets In

Connectors, ONTAP, and OCR for paper records.

3

Data: Real-time & Stored. From One Place.

Live vs. historical data, and why it matters.

4

Where Should Your AI Actually Run?

The deployment-posture brief.

5

Most AI Is Shared, Not Private

The shared-vs-private distinction, made concrete.

6

Where Regulation Makes Private AI the Safer Bet

The compliance case.

7

Five Doors Into Your AI’s Memory

The practitioner-level memory-risk paper.

8

The Puzzle Effect

How AI stitches together proprietary data over time.

9

You Already Own the Walls

The competitive-moat argument, one level deeper.

10

Where the Money Was Hiding

The $17.4M claims-recovery proof point.

11

Public, Private, Sovereign AI Compared

The three deployment postures, side by side: who controls it, what it costs, who owns what it learns.

12

The 5 Risk Vectors of AI Memory

The two-page companion to the Expert-tier deep dive on AI memory.

Deep Dive · Expert

13

What Is a Harness?

The difference between just chatting with AI and doing real work.

14

Inference: What Actually Runs the Model (Lifeboat)

The engine inside an AI that actually generates the responses.

15

LLM vs SLM

Large Language Model or Small Language Model, and why small is sometimes better.

16

Why Smaller Models Often Win

A smaller model designed for a specific task will perform better than a large general.

17

The Next Major Transition of Storage

Archives have evolved to become searchable resources.

18

From Stored to Answered

An AI platform to just get work done, no assembly required.

19

Four Myths About Private AI

Debunking the most common objections to running AI privately.

20

Hybrid Isn’t Indecision

The best tool for the job, and that often means more than one tool.

21

Six AI Risks Your IT Controls Can’t Catch

AI risks are different, and traditional IT might not work as before.

22

The One-Way Learning Loop

How you chat with an AI is just as important as the content.

23

When Your Prompts Become Permanent

What AI models remember about you is evolving rapidly.

24

“Your Data Doesn’t Train Our Model.” Technically.

Watch what your AI provider says and doesn’t say in the contract.

25

Patents: Purpose-Driven and Practical

Iterate.ai is building families of IP for private AI performance.

26

Protecting Institutional Knowledge in the AI Era — Deep Dive

Public AI can extract institutional knowledge in a slow invisible drip.

27

The End of “Per Token” Costs — Deep Dive

How Private AI changes the economics of adoption and success.

Take Certification Exam

To earn it: read all 27 briefs and pass a 40-question exam at 85%. Prerequisite: Basic.