DNAi Professional Certification

Before you build anything, you map how the work actually flows and decide where AI augments judgment versus where it safely automates rote steps. This module sets the operator's frame for every decision that follows.

• Map the Work Before the Tool
• Automate Versus Augment: The Decision Test
• Scoping the First Use Case for a Provable Win
• Building the Business Case Without Overpromising

Most useful AI inside a company is retrieval over your own knowledge, done so answers are grounded, traceable, and current. This module covers building retrieval that operators can trust.

• Why Retrieval Beats a Smarter Prompt
• Chunking, Citations, and Freshness
• Capturing the Knowledge in People's Heads

Agents chain tools and steps to carry a workflow further than a single prompt, which is powerful and easy to get dangerously wrong. This module covers scoping, guardrails, and knowing when an agent is the wrong answer.

• When You Actually Need an Agent
• Guardrails and Blast Radius
• Instrumenting for the Failure You Cannot See
• Defending Against Prompt Injection and Tool Abuse

The point of routing AI correctly is to reduce the mental tax on your people, not just to move numbers. This module makes cognitive load measurable and gives you levers to lower it without hiding work or eroding skill.

• Making Cognitive Load Measurable
• Lowering Load Without Hiding It
• Guarding Against Deskilling
• Designing the Interface Where Load Is Won or Lost

A rollout is where good intentions meet org reality, and where institutional knowledge and morale are most easily destroyed. This module is the operator's playbook for deploying AI without hollowing out the team.

• Sequencing a Rollout People Trust
• Honesty About Jobs and Morale
• Feedback Loops That Keep It Honest

An AI deployment is only as trustworthy as the way you measure it, and most operators measure the wrong things or stop measuring once the demo lands. This module builds the evaluation discipline that separates a system you can defend to a regulator, a board, and your own team from one that merely looks good on a slide. You learn to write an eval set that captures real-world variety, to choose metrics that resist gaming, and to keep the scorecard honest as the world drifts under the system.

• Writing an Eval Set That Reflects Reality
• Metrics That Resist Gaming
• Running the Eval Loop Before Users Do
• Reporting Results With Calibrated Honesty

A universal prompting checklist, then the specific habits that get the most out of Claude, ChatGPT, Gemini, and Perplexity.

• The universal checklist
• Prompting Claude
• Prompting ChatGPT and Gemini
• Prompting Perplexity

Match each job to the best tool, learn the vocabulary that specifies exactly what you want built, and chain tools into a simple workflow you can evaluate and improve.

• Match the tool to the job
• Speak in specs
• Chaining tools into a workflow
• Evaluate and iterate