Explain AI and agents without hype
What AI does well, where it fails, how much it costs and what already counts under the rules and LGPD, on one page for leadership.
The problem
Building has become easy. Trusting what was built hasn’t.
The company doesn't need one more builder. It needs someone who says what is worth it, what is a risk and what nobody has reviewed. Each problem below has a track that deals with it.
- Leadership wants AI and doesn't know which process to start with.
- AI-built apps connected to the CRM and the customer spreadsheet, with nobody reviewing keys, permissions or data.
- The manager knows the process by heart, but doesn't know how to turn it into a safe tool.
- The team took a prompt training, and on Monday went back to working as before.
- Nobody can say what AI does well, where it fails and how much it costs.
After the training
By the end of the Foundation and your track, you will be able to:
Diagnose and propose
Discovery conversation, value × feasibility, a 30-day pilot, measurement at 30, 60 and 90 days, and a priced proposal.
Audit what the company already uses
AI inventory with an owner, risk triage with evidence, app and agent review, and a report with a 30-day plan.
Mentor the manager who builds
Decide whether it's worth building, a one-day prototype, test before connecting, and a safe handoff to IT.
Train teams and executives
A 1-hour workshop, a 4-hour workshop or a bootcamp in the same structure, adapted to the learner's area and assessed by rubric.
5 courses in order · 40 lessons
One common foundation and four independent tracks.
The Common Foundation is the recommended starting point. After that, take one track or all four, in any order. In the company they complement each other: trainer in the workshop, consultant in the diagnosis, mentor in the 90 days and auditor in the risk review.
- 0
Shared foundationv7.2
Shared foundation
The 2027 map, from chat to agent, vibe coding from the trust side, the risks that sink projects, LGPD, scope and a page for leadership.
8 lessonsOpen the Common Foundation - 1
Track 1v7.2
AI Consultant
From the discovery conversation to the 30-day pilot with an agent, measurement at 30, 60 and 90 days, a priced proposal and the Horizonte diagnosis.
8 lessonsOpen track 1 - 2
Track 2v7.2
AI Auditor
Written authorization, AI inventory, risk triage, vibe audit, agent guard, data guard and the report with evidence and a 30-day plan.
8 lessonsOpen track 2 - 3
Track 3v7.2
AI Mentor
The manager who builds: is it worth building, a one-day prototype, autonomy with limits, tests, handoff to IT and the 90-day program.
8 lessonsOpen track 3 - 4
Track 4v7.2
AI Trainer
The 5 principles for teaching adults, from a 1-hour workshop to a bootcamp, delegating and reviewing instead of the magic prompt, and assessing the class.
8 lessonsOpen track 4
Who it's for
Made for you if
- You work with AI and want to be the trusted person the company calls, not just another builder.
- You come from IT, compliance or auditing and want to review what employees create with AI.
- You teach or train, and you want to build teams that use AI on Monday morning.
- You want a certification that clients can check on the spot, by code and QR.
Not for you yet if
- You are looking for pentesting, incident response or a legal opinion: those remain with qualified professionals.
- You want to sell consulting focused on prospecting and closing: start with the AI Consulting area.
Included in the training
You practice in a company that has deliberate flaws.
Horizonte Distribuidora is the program's fictional lab company, with 12 planted flaws for you to find. It comes with the area's projects and video.
Projects and kits
Videos
Why this training
What is different from a generic AI course.
One profession per track
Consultant, Auditor, Mentor and Trainer have separate courses, covering what each one does and who they work with.
Written limits
Each track says what you are not and when to call a specialist. The auditor does risk triage, not pentesting.
Practice in the lab company
The exercises use Horizonte Distribuidora and end in a deliverable: a diagnosis, a report, a program or a workshop kit.
Verifiable certification
Exam at cert.inema.pro and a certificate with code and QR, one per track. Rules and dates coming soon.
Questions
Before you start.
Do I need to take all four tracks?
No. Each track is independent. Take the Common Base first and then one, two or all four, in any order.
How long does it take?
Each course has 8 lessons. The Common Base plus one track add up to 16 lessons; the complete program, 40.
Is the certification available yet?
Not yet. Rules, prices and dates will be published soon on the area page. It is an INEMA qualification certification, not a regulated professional certification.
Are the courses available in English and Spanish?
Today they are in Portuguese. The English and Spanish versions will follow.
Do I need to know how to code?
No. The Common Base has a vibe coding lesson from the trust angle: you build a simple app and look for the flaws in it.
The work doesn't come from building. It comes from being someone people can trust when things get tight.
Start with the Common Base. In 8 lessons you learn what AI does well, where it fails and how much it costs, and you choose your track.