1:44Watch lesson →Visual lesson 01
What modern AI systems do
Tell which part of an AI system does each job.
Try the practice →See the idea. Try the skill.
AI lessons,
made visible.
Follow a source check, inspect a permission boundary, and see a flawed result become a useful one. Then practice independently in the linked lesson.
Illustrated examples use fictional data and simulated behavior. They are teaching companions, not provider benchmarks. Narration is synthesized locally; no AI account is needed to watch or use the supplied practice.
149 videos shown
No matching videos. Try another skill or choose all topics.
1:44Watch lesson →
1:40Watch lesson →Visual lesson 02
Safe use before prompting
Decide what information is safe to use before asking AI.
Try the practice →
1:47Watch lesson →Visual lesson 03
Prompt like an operator: Context, Task, Standards, Evidence
Give AI a clear task and check whether it followed your instructions.
Try the practice →
1:44Watch lesson →Visual lesson 04
Verification, sources, and calibrated confidence
Check important statements against the source.
Try the practice →
1:40Watch lesson →Visual lesson 05
Research, summarization, and drafting
Combine documents without hiding disagreements.
Try the practice →
1:47Watch lesson →Visual lesson 06
Decision support and human oversight
Use AI to compare options while a person owns the decision.
Try the practice →
1:47Watch lesson →Visual lesson 07
Tools, multimodal inputs, and agent boundaries
Keep information from a webpage from becoming permission to act.
Try the practice →
1:45Watch lesson →Visual lesson 08
Foundations field exercise
Create a clear instruction sheet using only supported steps.
Try the practice →
1:39Watch lesson →Visual lesson 09
Mission, roles, and the learning contract
Choose a learning path that fits the skills you can show.
Try the practice →
1:46Watch lesson →Visual lesson 10
Data handling and tool authorization gate
Check permission before sharing information with an AI tool.
Try the practice →
1:40Watch lesson →Visual lesson 11
Diagnostic and safe workspace setup
Practice allowed and blocked actions in a safe workspace.
Try the practice →
1:40Watch lesson →Visual lesson 12
Workflow audit and use-case selection
Choose a useful AI task and count the work it really saves.
Try the practice →
1:43Watch lesson →Visual lesson 13
Reusable CTSE workflows and quality standards
Turn a useful request into a workflow someone else can repeat.
Try the practice →
1:35Watch lesson →Visual lesson 14
Data preparation, provenance, and records
Prepare data without losing where it came from.
Try the practice →
1:44Watch lesson →Visual lesson 15
Source-grounded research and retrieval
Find sources that support the exact answer you need.
Try the practice →
1:41Watch lesson →Visual lesson 16
Documents, spreadsheets, and multimodal work
Check AI-created documents and charts against the original data.
Try the practice →
1:41Watch lesson →Visual lesson 17
Workflow evaluation and operational metrics
Test a workflow’s important failures, not just its average score.
Try the practice →
1:43Watch lesson →Visual lesson 18
Team rollout, change control, and incident handling
Roll out changes with an owner, a stop rule and a recovery plan.
Try the practice →
1:43Watch lesson →Visual lesson 19
Applied workflow capstone
Deliver a checked workflow and explain what it can and cannot do.
Try the practice →
1:42Watch lesson →Visual lesson 20
What AI actually is
Use AI to draft a useful answer, then check it against facts.
Try the practice →
1:46Watch lesson →Visual lesson 21
Five useful prompts and how to check them
Use five everyday AI tasks without changing the facts.
Try the practice →
1:44Watch lesson →Visual lesson 22
Prompt like an Operator
Write a clear request and improve it after checking the result.
Try the practice →
1:47Watch lesson →Visual lesson 23
When AI is wrong
Catch errors in meaning, numbers and approval claims.
Try the practice →
1:42Watch lesson →Visual lesson 24
AI for your transition
Use AI for career drafts without inventing experience.
Try the practice →
1:46Watch lesson →Visual lesson 25
AI vocabulary for choosing a workflow
Choose an AI setup by the jobs its parts actually do.
Try the practice →
1:37Watch lesson →Visual lesson 26
Where to go from here
Build a practice plan around the next skill you need.
Try the practice →
1:42Watch lesson →Visual lesson 27
Set up a workspace your AI can navigate
Organize files so an AI tool can find sources without getting extra access.
Try the practice →
1:39Watch lesson →Visual lesson 28
Turn a broad question into a research brief
Turn a broad topic into a question you can answer.
Try the practice →
1:40Watch lesson →Visual lesson 29
Build a source register and resolve conflicting versions
Choose the source that actually applies, not just the newest-looking page.
Try the practice →
1:40Watch lesson →Visual lesson 30
Construct a claim ledger before drafting
Match each important statement to evidence before drafting.
Try the practice →
1:39Watch lesson →Visual lesson 31
Write a briefing that separates findings from recommendations
Write a briefing with facts, advice and limits clearly separated.
Try the practice →
1:38Watch lesson →Visual lesson 32
Research across long documents without losing provenance
Keep exceptions when summarizing long documents.
Try the practice →
1:40Watch lesson →Visual lesson 33
Capstone: deliver a defensible decision briefing
Deliver a briefing someone can trace back to its sources.
Try the practice →
1:37Watch lesson →Visual lesson 34
Choose a useful task and define done
Define the output and the actions allowed before using AI.
Try the practice →
1:37Watch lesson →Visual lesson 35
Draft correspondence without changing commitments
Improve a message’s tone without adding a commitment.
Try the practice →
1:38Watch lesson →Visual lesson 36
Turn meeting notes into checked actions
Turn notes into actions without inventing agreement or owners.
Try the practice →
1:43Watch lesson →Visual lesson 37
Create procedures with exceptions and recovery
Write a procedure that handles missing inputs and exceptions.
Try the practice →
1:40Watch lesson →Visual lesson 38
Measure workflow improvement honestly
Measure the whole workflow rather than only faster drafting.
Try the practice →
1:36Watch lesson →Visual lesson 39
Capstone: assemble a checked work packet
Make related documents agree with the same source facts.
Try the practice →
1:39Watch lesson →Visual lesson 40
Define the question and the data contract
Define what you are counting before calculating an average.
Try the practice →
1:41Watch lesson →Visual lesson 41
Clean data without hiding your decisions
Clean data with a record of what you changed and why.
Try the practice →
1:41Watch lesson →Visual lesson 42
Verify calculations with executable evidence
Distinguish percentage-point change from relative change.
Try the practice →
1:39Watch lesson →Visual lesson 43
Compare alternatives without overstating causation
Compare options using usable output and repair effort.
Try the practice →
1:41Watch lesson →Visual lesson 44
Communicate uncertainty and decision sensitivity
Show when uncertain assumptions could change the decision.
Try the practice →
1:42Watch lesson →Visual lesson 45
Capstone: produce a reproducible decision analysis
Deliver calculations another person can repeat and inspect.
Try the practice →
1:42Watch lesson →Visual lesson 46
Command line, Git, Python, and environments
Make your program run from a clean copy, not just your own folder.
Try the practice →
1:44Watch lesson →Visual lesson 47
HTTP, JSON, APIs, secrets, and failure handling
Handle service failures with clear retry and stop rules.
Try the practice →
1:43Watch lesson →Visual lesson 48
Provider-neutral generation APIs
Connect different model services without hiding their differences.
Try the practice →
1:42Watch lesson →Visual lesson 49
Data engineering for AI applications
Prepare mixed data with rules you can repeat and inspect.
Try the practice →
1:41Watch lesson →Visual lesson 50
Software quality, testing, and safe telemetry
Test your software without copying secrets into logs.
Try the practice →
1:40Watch lesson →Visual lesson 51
Structured outputs and bounded tools
Check structured answers before allowing a tool to run.
Try the practice →
1:38Watch lesson →Visual lesson 52
Retrieval baseline from first principles
Build a small search test before adding more retrieval machinery.
Try the practice →
1:42Watch lesson →Visual lesson 53
Production RAG: authorization, hybrid search, and grounding
Retrieve only information the user is allowed to receive.
Try the practice →
1:42Watch lesson →Visual lesson 54
Evaluation design and release gates
Set test requirements before deciding whether to release.
Try the practice →
1:44Watch lesson →Visual lesson 55
Agents and tool use under control
Control an agent’s actions and keep its state outside the conversation.
Try the practice →
1:42Watch lesson →Visual lesson 56
Deployment, operations, and cost evidence
Check readiness and recovery before promoting a new version.
Try the practice →
1:38Watch lesson →Visual lesson 57
Builder capstone and technical defense
Show a small application’s behavior and defend its limits.
Try the practice →
1:43Watch lesson →Visual lesson 58
Machine learning foundations for system builders
Check whether a model’s confidence matches its observed results.
Try the practice →
1:45Watch lesson →Visual lesson 59
Transformer and inference internals
Explain why longer inputs and outputs can change serving cost and timing.
Try the practice →
1:42Watch lesson →Visual lesson 60
Adaptation, fine-tuning, and post-training decisions
Test whether adaptation fixes the needed behavior without breaking another.
Try the practice →
1:41Watch lesson →Visual lesson 61
Production architecture and inference serving
Test service capacity under bursts, not just average demand.
Try the practice →
1:42Watch lesson →Visual lesson 62
AI threat modeling and adversarial testing
Find and block the action an attacker is trying to cause.
Try the practice →
1:40Watch lesson →Visual lesson 63
Identity, authorization, privacy, and data protection
Enforce the same access rule on every route to data.
Try the practice →
1:40Watch lesson →Visual lesson 64
Evaluation engineering and statistical evidence
Compare systems without averaging away serious failures.
Try the practice →
1:42Watch lesson →Visual lesson 65
Observability, safe logging, and incident response
Contain a failing route and prove that recovery works.
Try the practice →
1:42Watch lesson →Visual lesson 66
Reliable workflows and agents
Recover repeated work without causing duplicate effects.
Try the practice →
1:43Watch lesson →Visual lesson 67
Performance, reliability, and resource governance
Improve performance without silently changing the task.
Try the practice →
1:38Watch lesson →Visual lesson 68
Secure supply chain and continuous delivery
Know which software, models and data are in your release.
Try the practice →
1:44Watch lesson →Visual lesson 69
Production capstone, incident exercise, and board
Show that your training service can handle a failure and recover.
Try the practice →
1:41Watch lesson →Visual lesson 70
ML and deep learning: enough to be dangerous
Judge a classifier by its errors as well as its accuracy.
Try the practice →
2:16Watch lesson →Visual lesson 71
LLM internals: how transformers work
Explain how a text model processes pieces of a prompt.
Try the practice →
1:48Watch lesson →Visual lesson 72
Fine-tuning, post-training, and alignment
Choose an AI improvement based on the failure you observed.
Try the practice →
1:47Watch lesson →Visual lesson 73
Production deployment patterns and infrastructure
Plan a service around real workload and required behavior.
Try the practice →
1:42Watch lesson →Visual lesson 74
Monitoring, observability, and incident response
Find out what an alert means and verify recovery after a failure.
Try the practice →
1:43Watch lesson →Visual lesson 75
AI security: prompt injection, model risk, supply chain
Test whether forbidden actions are blocked, not just whether AI says no.
Try the practice →
1:45Watch lesson →Visual lesson 76
Federal AI compliance: NIST AI RMF from the engineer's seat
Connect AI risks to evidence, owners and decisions.
Try the practice →
1:44Watch lesson →Visual lesson 77
ATO and FedRAMP for AI systems
Check the whole system’s authorization boundary.
Try the practice →
1:40Watch lesson →Visual lesson 78
Procurement from the engineer's chair
Translate proposal requirements into evidence you can deliver.
Try the practice →
1:44Watch lesson →Visual lesson 79
Capstone: production AI in a federal context
Defend a complete training project using reproducible evidence.
Try the practice →
1:38Watch lesson →Visual lesson 80
Python for AI work
Write a small Python program that rejects invalid input.
Try the practice →
1:42Watch lesson →Visual lesson 81
OpenAI and Anthropic APIs, end to end
Handle an AI API response without mistaking partial text for success.
Try the practice →
1:38Watch lesson →
1:39Watch lesson →Visual lesson 83
Prompt engineering at scale
Reuse prompts while checking both output format and meaning.
Try the practice →
1:44Watch lesson →Visual lesson 84
Retrieval-augmented generation, from scratch
Give an AI answer relevant evidence and check whether it used it.
Try the practice →
1:42Watch lesson →Visual lesson 85
Vector databases and embeddings
Keep document and query embeddings compatible.
Try the practice →
1:40Watch lesson →Visual lesson 86
Agents and tool use
Let an agent choose tools without letting it grant itself access.
Try the practice →
1:39Watch lesson →Visual lesson 87
Evaluation and observability
Test whether an AI result meets the task, not just whether it contains a keyword.
Try the practice →
1:42Watch lesson →Visual lesson 88
Cost, latency, and trade-offs
Measure response time and usage without hiding retries or unknowns.
Try the practice →
1:39Watch lesson →Visual lesson 89
Capstone: ship something real
Build one small application that another person can inspect.
Try the practice →
1:42Watch lesson →Visual lesson 90
Specify an application before choosing a model
Specify the application’s job before choosing a model.
Try the practice →
1:38Watch lesson →Visual lesson 91
Give coding assistants bounded context and inspect the diff
Give a coding assistant a narrow change and inspect what it actually changed.
Try the practice →
1:36Watch lesson →Visual lesson 92
Handle structured output and provider failure states
Check both the shape and the truth of a structured result.
Try the practice →
1:35Watch lesson →Visual lesson 93
Enforce tool permissions and preserve denial
Preserve a denied read instead of pretending the record was inspected.
Try the practice →
1:39Watch lesson →Visual lesson 94
Evaluate failures and bound repeated work
Bound repeated work and describe what the teaching cache does not prove.
Try the practice →
1:46Watch lesson →Visual lesson 95
Capstone: deliver a bounded offline workflow
Deliver an offline workflow with honest execution evidence.
Try the practice →
1:40Watch lesson →Visual lesson 96
Role families and evidence-based gap analysis
Match your demonstrated work to the tasks in a job posting.
Try the practice →
1:36Watch lesson →Visual lesson 97
Resume, profile, and interview translation with OPSEC
Write career claims that accurately describe your contribution.
Try the practice →
1:44Watch lesson →Visual lesson 98
Portfolio, technical narrative, and credential selection
Make a portfolio repeatable and safe to share.
Try the practice →
1:41Watch lesson →Visual lesson 99
Career campaign and evidence defense
Explain a project’s result and limits in an interview.
Try the practice →
1:47Watch lesson →Visual lesson 100
Mission, information, and authorization boundaries
Separate public study material from information needing an approval decision.
Try the practice →
1:45Watch lesson →Visual lesson 101
OPSEC and approved AI use
Check whether combined details reveal more than each detail alone.
Try the practice →
1:45Watch lesson →Visual lesson 102
FedRAMP 20x, cloud impact levels, and authorization boundaries
Check which exact services a cloud assessment covers.
Try the practice →
1:43Watch lesson →Visual lesson 103
RMF and the Authority to Operate
Show that an access control works before calling it implemented.
Try the practice →
1:40Watch lesson →Visual lesson 104
Continuous authorization and DevSecOps evidence
Retest the workflow when a component or data path changes.
Try the practice →
1:46Watch lesson →Visual lesson 105
Federal AI governance and high-impact use
Judge an AI feature by what it does and who it affects.
Try the practice →
1:41Watch lesson →Visual lesson 106
On-premises, disconnected, and open-weights operations
Trace every data path before claiming a setup is local-only.
Try the practice →
1:41Watch lesson →Visual lesson 107
Federal acquisition artifacts for engineers
Write technical requirements that someone can actually test.
Try the practice →
1:39Watch lesson →Visual lesson 108
Federal specialization capstone and board
Reopen permissions when a project expands from drafting to acting.
Try the practice →
1:35Watch lesson →Visual lesson 109
Leader's model of AI capability and accountability
Make human review a real check, not just a label.
Try the practice →
1:39Watch lesson →Visual lesson 110
Use-case portfolio, value, and stop decisions
Compare the whole task with AI and without it.
Try the practice →
1:41Watch lesson →Visual lesson 111
Vendor evaluation and mission bake-off
Evaluate vendors using the same task and predeclared rules.
Try the practice →
1:43Watch lesson →Visual lesson 112
Requirements, test strategy, and acceptance
Replace “be accurate” with a requirement two reviewers can test.
Try the practice →
1:37Watch lesson →Visual lesson 113
AI risk governance and incident leadership
Keep a failing feature paused until each important problem is repaired.
Try the practice →
1:45Watch lesson →Visual lesson 114
Funding, lifecycle resources, and fiscal role discipline
Estimate workload, review time and retries before planning resources.
Try the practice →
1:38Watch lesson →Visual lesson 115
Team, operating model, and 180-day leadership practicum
Scale only when staffing and recovery evidence support the plan.
Try the practice →
1:39Watch lesson →Visual lesson 116
What cleared AI work actually means
Separate AI skills from clearance and access requirements.
Try the practice →
1:42Watch lesson →Visual lesson 117
OPSEC for AI tools
Check the information in a request, not just the topic of its answer.
Try the practice →
1:40Watch lesson →Visual lesson 118
On-prem and air-gapped AI
Check whether every component stays inside the intended boundary.
Try the practice →
1:37Watch lesson →Visual lesson 119
The cleared AI role landscape
Choose roles by their work, not by a shared AI title.
Try the practice →
1:44Watch lesson →Visual lesson 120
The certs that count
Check what a credential demonstrates before relying on it.
Try the practice →
1:42Watch lesson →Visual lesson 121
Talking about your work without violating
Describe work truthfully and within its release rules.
Try the practice →
1:44Watch lesson →Visual lesson 122
Where to go from here
Advance learning while waiting for required tool approval.
Try the practice →
1:43Watch lesson →Visual lesson 123
What you actually need to know about AI as a leader
Ask what the whole AI system can do, not just which model it uses.
Try the practice →
1:35Watch lesson →Visual lesson 124
Vendor evaluation that doesn't get gamed
Choose candidates using critical rules before average scores.
Try the practice →
1:44Watch lesson →Visual lesson 125
Writing AI requirements into a contract
Write requirements with a source, a check and a decision owner.
Try the practice →
1:42Watch lesson →Visual lesson 126
Risk management at the leader's altitude
Tie each risk to a particular feature and consequence.
Try the practice →
1:45Watch lesson →Visual lesson 127
The budget cycle and color of money
Estimate workflow demand without treating the estimate as spending approval.
Try the practice →
1:40Watch lesson →Visual lesson 128
Building and managing an AI team
Build backup coverage that someone has actually demonstrated.
Try the practice →
1:41Watch lesson →Visual lesson 129
The first 180 days as an AI leader
Use checkpoints to decide when scaling is justified.
Try the practice →
1:33Watch lesson →Visual lesson 130
Select a use case with a real decision and baseline
Start with the work problem rather than a product purchase.
Try the practice →
1:41Watch lesson →Visual lesson 131
Understand the complete system boundary
Ask where data goes and which actions the whole system allows.
Try the practice →
1:41Watch lesson →Visual lesson 132
Evaluate vendor claims with task-specific evidence
Ask what a vendor’s headline score actually measured.
Try the practice →
1:36Watch lesson →Visual lesson 133
Design a bounded pilot with stop conditions
Plan a small trial with checks that can stop it.
Try the practice →
1:40Watch lesson →Visual lesson 134
Plan adoption, changes and failure recovery
Keep ownership and testing current after the first demonstration.
Try the practice →
1:41Watch lesson →Visual lesson 135
Capstone: write an accountable AI adoption decision
Write a conditional adoption decision that shows what is still missing.
Try the practice →
2:48Watch lesson →Visual lesson 136
Your first useful AI result
Repair an unsupported draft using supplied evidence.
Try the practice →
2:34Watch lesson →Visual lesson 137
What are you actually using?
Distinguish the model, application, harness and execution environment.
Try the practice →
2:30Watch lesson →Visual lesson 138
Turn a vague request into a usable task
Write and evaluate a Context, Task, Standards and Evidence prompt.
Try the practice →
2:20Watch lesson →Visual lesson 139
Local, cloud and frontier: separate the questions
Trace deployment, capabilities and complete data destinations.
Try the practice →
2:33Watch lesson →Visual lesson 140
Same model, different harness
Explain how tools, permissions and checks change a workflow.
Try the practice →
2:20Watch lesson →Visual lesson 141
Context, memory and retrieval
Recover from stale summaries and missing source context.
Try the practice →
2:36Watch lesson →Visual lesson 142
Verify an answer before using it
Separate supported facts, inference, quotations and action claims.
Try the practice →
2:34Watch lesson →Visual lesson 143
An agent requests access—and gets denied
Interpret permission denial and bounded tool execution honestly.
Try the practice →
2:04Watch lesson →Visual lesson 144
Valid JSON can still be wrong
Validate output structure, factual values and failure states separately.
Try the practice →
2:09Watch lesson →Visual lesson 145
Retrieval that answers—or abstains
Check retrieved sources and decline unsupported answers.
Try the practice →
2:34Watch lesson →Visual lesson 146
Claimed completion versus tested completion
Verify actual artifacts and requirements before reporting success.
Try the practice →
2:12Watch lesson →Visual lesson 147
Build evidence of your capability
Describe your contribution, AI assistance and limitations truthfully.
Try the practice →
10:36Watch lesson →Visual lesson 148
What AI actually is
Explain the difference between a generated claim and supporting evidence. Compare two prompts against the same source facts. Repair a draft that invents a fact, then explain the correction.
Try the practice →
11:41Watch lesson →Visual lesson 149
AI API requests and safe response handling
Translate a supported synthetic request and inspect text, usage and completion status. Distinguish simulated permanent and transient failures, and state live recovery limits. Record the actual runtime, fixture revision and verification result; identify the SDK/model evidence required for live reproduction.
Try the practice →