Wirewalk Instituteof AI
pub-220 Intermediate AvailableFlagship

Building and Deploying AI in Defense: Logistics, Maintenance, Intelligence Analysis and Training

For engineers, data scientists, test engineers and assurance leads in the Department of Defense and the defense industrial base who must build, test and field AI for specific maintenance, supply, analytic-support and training tasks under current DoD, CMMC and export rules.

Level
Intermediate
Study time
16 hours
Lessons
19 in 6 modules
Exam
20 questions

About this course

A 16-hour planned supplied-output course for program assurance, sustainment and contractor technical teams. Preserve established maintenance, acquisition and instructor authority while evaluating where AI can help. Work through discrepancy coding,40-tail drift checks, current/effectivity-aware manual retrieval, intermittent demand and PR exceptions, PAI provenance/access and cited training AARs. A cumulative Pine Harbor case leads to a costed continue/narrow/hold memorandum.

November 14, 2026 author refresh distinguishes primary-source findings from blocked or unverified current documents. DoD’s July 2026 Phase II suspension announcement is separated from the published CMMC schedule and actual contract requirements. Sources and limitations are recorded in the source ledger; no provider, deployment or legal approval is supplied.16 hours is an author planning estimate, not measured learner timing; code/live-platform extensions add time. AI practice feedback is fallible and advisory; its score is not a certificate gate. Completion follows lessons, seven listed labs and the exam; it does not certify workplace competence, vendor endorsement or legal compliance.

What you will be able to do

  • Create a policy applicability register separating cited baseline, current-source gaps, binding obligations, policy announcements and local controls, with owners and recheck triggers
  • Specify a CBM+ failure-risk deployment for a synthetic fleet, including the model card's out-of-scope uses, drift and out-of-distribution checks, a concrete time-based fallback trigger, and success measures such as alert precision and lead time
  • Specify and evaluate a technical-manual retrieval assistant using synthetic citation, current-revision, supplement, effectivity and verbatim-safety tests, including correct abstention and the limits of numeric/citation screens
  • Create a component data-flow boundary and two focused SSP control paragraphs from supplied CDI/ITARrecords,with permission/security gaps,authorized citizenship-or-residence determinations and incident handoffs explicit
  • Select an intermittent-demand method from supplied rolling-origin MASE and lead-time bias, define a fair comparison for an unvalidated ML candidate, and write a forecast exception/readiness handoff that preserves purchase authority and missing-data limits
  • Audit a synthetic PAI source pipeline for duplicate inflation, translation/provenance and current access, and separate source support, analyst judgment and the authorities governing the cell’s actual activity
  • Create a compact decision-linked CBM+test record with scope,mature-label/split rules,two evaluation decisions,monitoring owner and fallback/resume conditions
  • Draft an after-action review from a synthetic convoy-resupply exercise log in which every finding cites a log line

Syllabus

6 modules · 19 lessons · about 16 hours

Module 1: Before You Build: The Assurance Lead's Applicability Matrix and the Contractor's CUI/ITAR Boundary

An assurance lead decides which defense AI policies apply to five fictional project briefs, and a contractor platform engineer scopes an AI assistant for CUI and ITAR data at the fictional Synthetic Gearworks LLC.
4 lessons
  1. Five Project Briefs, One Matrix: DoD AI Ethical Principles, RAI Toolkit SHIELD and DAGR, the January 2026 AI Strategy, NSPM-11 and NATO Principles
  2. DoDD 3000.09 Scope for Builders: Why an Unarmed Depot Tug Is Outside It and a Future Armament Option Triggers Review
  3. An AI Assistant for Covered Defense Information: DFARS 252.204-7012, FedRAMP Moderate Equivalency, 72-Hour Reporting and CMMC Phase 2
  4. ITAR Technical Data in Prompts and Indexes: 22 CFR 120.54(a)(5), Foreign-Person Access and the SSP Rewrite for Synthetic Gearworks LLC
  • Decision Lab · 12 minThe Supplier Boundary Before the DemonstrationDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Module 2: Aircraft Maintenance: A Sustainment Data Scientist Fielding Discrepancy Coding and CBM+ Risk Alerts

Clean free-text discrepancy narratives, put a CBM+ failure-risk model into maintenance planning without ever extending an inspection interval, and write the T&E plan that clears it for use.
4 lessons
  1. Coding Free-Text Discrepancy Narratives Into Failure and Action Codes Through a Review Queue, Scored per Class
  2. Putting a CBM+ Failure-Risk Model Into Maintenance Planning: IMDS, REMIS, NALCOMIS and GCSS-Army Data, PANDA-Style Systems and DoDI 4151.22
  3. Catching Sensor Drift and an Out-of-Distribution Airframe in a 40-Tail Synthetic Fleet, and Writing the Time-Based Fallback Trigger
  4. Writing the T&E Plan for the CBM+ Model: DoDI 5000.98, DoDM 5000.101, Time-Based Test Splits, Monitoring and Rollback
  • Decision Lab · 12 minEight Review Slots and Three Different ExceptionsDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Module 3: Maintainers' Technical-Manual Assistant: Retrieval That Never Drops a WARNING

A software team builds and evaluates a retrieval assistant over technical manuals that cites the exact paragraph and revision, keeps safety text verbatim and respects distribution statements.
3 lessons
  1. Indexing TMs and IETMs Under DoDI 5230.24 Distribution Statements, CUI Markings, ITAR Controls and Cloud Impact Levels
  2. Verbatim WARNING and CAUTION Text, Current-Revision Retrieval and the "Not in the Manual; Contact Engineering" Answer
  3. Scoring the Assistant on TM 99-XYZ-SYNTH: Citation, Revision and Warning-Preservation Tests Where DoDM 5000.101 Does Not Reach Generative AI
  • Decision Lab · 12 minA Current Citation for the Wrong GeneratorDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Module 4: Supply and Readiness: A DLA Demand Planner and a Readiness Analyst Forecasting What Leaders Act On

Forecast intermittent spare-parts demand that planners can trust and explain, and forecast mission-capable rates on an enterprise data platform without presenting them as a readiness report of record.
3 lessons
  1. Backtesting Intermittent Spare-Parts Demand: Naive, Croston-Type and Machine-Learning Forecasts Scored With MASE
  2. Explaining Forecast Changes to the Planner and Screening Them Before Automated Purchase Requests
  3. Forecasting Mission-Capable Rates on Army Vantage or Advana: Drivers, Reporting Lag and a One-Slide Commander Brief
  • Decision Lab · 12 minThe Lowest Error and the Wrong ReplenishmentDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Module 5: Analytic-Cell Engineer: PAI Triage and Summarization Over Finished Reporting

Build intelligence-analysis support tools that collapse duplicates, keep provenance, follow ICD 203 and ICD 206, and keep access labels intact on a classified network, using only synthetic unclassified data.
3 lessons
  1. Building a PAI Triage Pipeline for a Fictional Port Strike: Translation, Deduplication, Event Clustering and Provenance
  2. ICD 203 Likelihood and Confidence Rules, ICD 206 Sourcing and U.S.-Person Rules as Acceptance Criteria for AI-Drafted Text
  3. Summarization on a Classified Network, Taught With a Synthetic Access-Labeled Corpus: Index-Time Filtering, Caveats and Compilation Risk
  • Decision Lab · 12 minThree Copies and a Revoked PermissionDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Module 6: Simulation Center Training Developer: AI Scenario Injects and After-Action Reviews

Generate varied sustainment, convoy-movement and HADR training injects and draft after-action reviews in which every finding traces to the exercise log, with instructors keeping certification decisions.
2 lessons
  1. Generating Sustainment, Convoy-Movement and HADR Scenario Injects and AI Role Players for VBS4 and Other Virtual Training Environments
  2. Drafting After-Action Reviews That Cite Every Finding to a Log Line, With DoDI 1322.26 Learning Records and Privacy Act Controls
  • Decision Lab · 12 minThe AAR Before the Next IterationDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.
  • Capstone simulation · 30 minThe December Board: A Demonstration Is Not FieldingDefend a bounded evidence-work scope for the Pine Harbor December demonstration; compare resource and reliance consequences.

Exam and certificate

The exam unlocks once you have completed every lesson and played each Decision Lab to an ending. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 40 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.

Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.

The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by C3 AI, Virtualitics, Palantir, Ask Sage, Microsoft, Google, OpenAI, xAI, Babel Street, Dataminr, Bohemia Interactive Simulations, Anthropic, or Amazon Web Services. Product and company names are trademarks of their respective owners.

More in AI in Government, Defense and Public Safety

All 12 courses in this track
pub-001 Free intro Available

AI in Government, Defense and Public Safety: The One-Hour Briefing

A free one-hour briefing built on twelve real public-sector jobs, not an overview: summarizing public comments in an approved assistant, extracting SNAP verification documents, flagging PII for FOIA, reviewing an AI-drafted police report, escalating from a 911 virtual agent, handling a facial-recognition lead, checking an AI bench memo, drafting on GenAI.mil, acting on a predictive-maintenance alert, and drafting a grant. Each lesson shows where AI helps, where it must not decide, and the rule that governs that exact task.

1 h · 9 lessons Free intro
pub-110 Basic Available

AI for Public Servants

Use approved AI for public-service drafting and evidence work while keeping decisions, privacy, records and review authority explicit. Preserve and correct concrete source artifacts, carry six module outputs into a synthetic pilot recommendation, and practice trade-offs in six module Decision Labs plus a capstone simulation.

8 h · 20 lessons See price at checkout
pub-210 Intermediate Waitlist

Building and Deploying AI in Government

How public bodies adopt AI as of October 2026: OMB M-25-21 and M-25-22 in practice, use case inventories, buying through GSA schedules, OneGov and FedRAMP 20x, service delivery use cases, due process in automated eligibility decisions, records and security obligations, and lessons from international public-sector frameworks.

12 h · 18 lessons Join the waitlist