AI foundations
Build a practical mental model of models, prompts, context, and uncertainty.
You will: Explain the difference between generating a useful answer and establishing that it is true.
AI FoundationsKNOWLEDGE LIBRARY
Find the idea.
Build the skill.
19 guides and 3 worked cases, organized into six connected domains.
Prefer a guided sequence? Choose your learning path →Explore by your next question
22 learning resources
No matching resources. Try fewer words or reset your filters.
Verify AI outputs
Trace claims to evidence and distinguish a real citation from actual support.
You will: Build a small evidence record and decide whether a claim is supported, limited, or unresolved.
Prompting & Practical AIReliable AI workflows
Define inputs, decision points, human review, and recovery before adding autonomy.
You will: Sketch a bounded workflow with visible checks and an explicit response to failure.
AI Systems & ArchitectureEvaluation & risk
Turn “looks good” into explicit criteria, representative cases, and release decisions.
You will: Create a small evaluation plan that exposes failures and supports a clear release decision.
Evaluation, Security & GovernanceBetter prompts
Turn vague requests into clear instructions, useful context, and checkable results.
You will: Write a reusable prompt and compare its results against a small checklist.
Prompting & Practical AIRAG explained
Understand retrieval-augmented generation and identify whether retrieval or the answer failed.
You will: Explain a RAG pipeline and diagnose a missing, outdated, or unsupported answer.
AI Systems & ArchitectureAI agents & permissions
Separate reasoning, tool use, and authority when designing an AI agent.
You will: Define allowed actions, approval points, and stop conditions for a bounded agent.
AI Systems & ArchitectureAI data privacy
Reduce unnecessary data exposure and ask better questions about AI services and connected tools.
You will: Prepare a minimal task input and map where sensitive information could travel.
Evaluation, Security & GovernanceBuild an AI evaluation
Build a small evaluation set, define observable checks, and compare changes without hiding critical failures.
You will: Create an evaluation record with cases, expected behavior, separate scores, and a justified release decision.
Evaluation, Security & GovernanceDiagnose RAG failures
Trace a wrong RAG answer through source availability, retrieval, context, and generation before choosing a repair.
You will: Produce a failure diagnosis supported by a source passage, retrieved results, and a targeted retest.
AI Systems & ArchitectureDesign meaningful human review
Define what reviewers inspect, what they can stop, and how errors reach a named owner.
You will: Write a review protocol with evidence, decision rights, escalation rules, and a tested fallback.
Evaluation, Security & GovernanceChoose an AI use case
Compare AI assistance, fixed rules, and a manual process before committing to a pilot.
You will: Write a one-page use-case brief with a baseline, boundaries, evidence requirements, and a clear next decision.
AI Strategy & ImplementationMeasure workflow value
Compare total effort, accepted output, quality, and ongoing overhead using a transparent pilot record.
You will: Calculate a fictional workflow comparison and explain which benefits are measured, assumed, or still unknown.
AI Strategy & ImplementationRespond to AI incidents
Design an operational response for incorrect outputs, unintended actions, and failures discovered after release.
You will: Prepare an incident runbook with containment, evidence, ownership, recovery, and restart checks.
Evaluation, Security & GovernanceConfidence & uncertainty
Separate fluent language, stated confidence, and verified support when deciding whether to use an AI answer.
You will: Label a claim as supported, unresolved, or contradicted and choose the next verification step.
AI FoundationsPrompting, RAG & fine-tuning
Distinguish task instructions, retrieved evidence, and trained behavior before adding complexity to an AI system.
You will: Diagnose a failure and justify whether to improve the prompt, evidence pipeline, training data, or ordinary application controls.
AI Systems & ArchitectureCompare AI models fairly
Compare candidates on matched tasks, repeated trials, review effort, and critical constraints rather than a single showcase answer.
You will: Write a model comparison memo with a controlled test, task-level results, trade-offs, and a bounded recommendation.
AI Systems & ArchitectureTraining & inference
Separate training, inference, and retrieval, then see why familiar examples are not enough to prove useful learning.
You will: Explain which part of an AI system changes during training and design a small test with unseen examples.
AI FoundationsVerify multimodal inputs
Check image and document interpretations by separating visible evidence, extracted text, and conclusions.
You will: Build a verification record for a chart or document image without turning unclear pixels into confident facts.
Advanced & Emerging AIThe convincing brief with unsupported claims
Audit an AI-assisted briefing, separate citation identity from claim support, and produce a defensible revision.
You will: A corrected brief and a claim-by-claim evidence record.
Prompting & Practical AIThe automation that moved the bottleneck
Follow a support-drafting pilot through misleading time savings, a failed handoff, and a bounded redesign.
You will: A workflow playbook with a baseline, approval rules, and recovery behavior.
AI Strategy & ImplementationThe agent that passed the wrong test
Inspect an apparently strong evaluation, uncover a critical permission failure, and write a defensible release memo.
You will: A release memo that separates average performance from critical control failures.
Evaluation, Security & Governance