Instructional designer · Learning experience designer · Available for contract

Training that changes what people do on Monday.

I take complicated subject matter and turn it into scenarios, simulations, and blended programs that hold up in the real job. Everything below is live, so click through it.

Behavior changeBranching scenarios with consequences, not 60-slide click-throughs.
Systems adoptionShow / Try / Test simulations with smart hints and job aids.
Scalable deliveryResponsive, blended programs that connect async, live, and on-the-job learning.
01 · Flagship branching scenario

The Change Request

A business email compromise scenario for accounts payable teams. Learners spot red flags, then make three decisions under pressure while a live risk meter reacts to every choice.

Compliance & riskBranching logicVariables & conditionsConsequence feedback
ClientMeridian Health Partners (fictional)
AudienceAP specialists & finance ops, ~120 learners
My roleAnalysis, scenario writing, UX, build
Seat time8–12 minutes, replayable
Business problem

The annual phishing course had a 98% completion rate, and the fraud still got through.

Two payment-redirection attempts reached the AP team in one quarter. One was paid. The existing training was a slide-based course that taught learners to recognize phishing but never asked them to act under the pressure attackers create: urgency, seniority, and secrecy.

The root cause wasn't knowledge. Everyone knew the callback policy. What failed was applying it when a VP was asking you to skip it.

Action map

Four observable behaviors, agreed with the controller and InfoSec.

Business goalZero unverified vendor banking changes within two quarters
  1. Spot the red flags in a payment-change requestLookalike domain, urgency, secrecy, bank change, odd attachment
  2. Verify through an independent, known-good channelVendor master phone number, never contact details from the email
  3. Hold firm against authority pressure and escalateCite policy, involve your manager and InfoSec
  4. Report and contain without destroying evidencePhish Alert button plus a payment hold. Don't delete.
Design decisions

Realistic pressure, honest consequences, no scolding.

Consequences come before feedback. Each choice shows what happens next in the story first, then explains why. Learners feel the cost before they hear the principle.

The "almost right" options are the most valuable. Replying to the email to "verify" feels responsible. It's also the most common real-world mistake, so it's a distractor on purpose.

The story remembers. Dana's pressure call changes depending on whether you already called the vendor. Choices carry forward instead of resetting at each decision.

Technical build

Same architecture as a Storyline build: variables, triggers, layers, states.

// Decision 2: conditional dialogue + variable adjustment Show state DanaLine when timeline starts if Verified == True → "I heard you called the vendor…" elif Choice1 == "A" → "Did the update go through?" else → "I confirmed by email…" Adjust RiskExposure += 30 when user clicks choiceA Show layer Outcome_{Stopped | CloseCall | Lost} // RiskExposure ≤15 | ≤50 | >50
7 tracked variablesNumber, true/false, and text variables drive the meter, the branching dialogue, and the personalized recap.
Hotspot interaction with statesRed-flag discovery uses Normal / Hover / Flagged states, and Next stays locked until FlagsFound ≥ 3.
Designer viewA live panel shows every variable and trigger as it fires, so reviewers can check the logic, not just the surface.
Web-nativeResponsive player, keyboard navigable, loads instantly. No plugin or LMS needed to review.
Measurement plan

What we'd track after launch

100%of vendor banking changes with a logged callback
↑ 3×Phish Alert reports from finance in the 90 days after launch
< 15average final risk score across learners' first attempt

Concept project. The client, people, and figures are fictional; targets show how I'd tie the design to business metrics.

02 · Systems simulation + job aid

Summit CRM: Update an Opportunity

A software simulation for a CRM rollout. Learners watch, practice with escalating hints, then prove it with no help. A one-page quick reference supports them on the job.

Software adoptionShow / Try / TestError-handling logicPerformance support
ClientSummit CRM rollout (fictional)
AudienceAccount executives, ~300 seats
My roleTask analysis, sim design, job aid, build
Seat time5–7 minutes across 3 modes
Business problem

Three weeks after go-live, 40% of opportunities still sat in the wrong stage.

Sales leadership couldn't trust the forecast. The vendor's training videos showed every feature; reps needed one workflow done correctly after every customer call. The task had to become a habit before the next quarterly forecast.

Design decisions

One workflow, practiced three ways.

Show me → Try me → Test me. A gradual release: a captioned demo, guided practice, then an unassisted assessment with an 80% pass mark.

Hints that escalate. The first wrong click in Try mode gets a specific text hint. The second also highlights the target. Learners get enough help to stay unstuck without being given the answer.

A job aid instead of more training. Choosing the right stage is a judgment call, not a click path, so it lives on the quick-reference card where reps can check it during real work.

Technical build

A data-driven step engine, not 24 hand-built slides.

// Each step is data: target, action type, validation, hint { t: "f-amount", type: "input", match: v => v.replace(/[$,\s]/g, "") === "48500", // accepts 48500, 48,500, $48,500 hint: "Click into the Amount field, type 48500, then press Enter." } on wrong click → WrongClicks += 1 if Mode == "try" → show hint; if misses ≥ 2 → pulse target if Mode == "test" and misses ≥ 2 → mark step failed, auto-complete
Forgiving validationAccepts the formats real people type ($48,500 / 48500) instead of failing them on punctuation.
Scripted demo cursorShow mode animates a cursor along each target with typed input, generated from the same step data.
Per-step resultsThe results screen shows which steps were missed, so learners know exactly what to practice.
Print-ready job aidUS Letter layout with print CSS. Saves to PDF in one click and reflows on a phone.
Measurement plan

What we'd track after launch

< 10%opportunities in the wrong stage at next forecast
−50%"how do I…" CRM help tickets in 30 days
80%+first-attempt pass rate in Test mode

Concept project. Summit CRM is a fictional product built for this simulation.

03 · Blended onboarding

Shift Lead Ready

A responsive, Rise-style onboarding module for newly promoted warehouse shift leads. It's step one of a blended path that continues into a live virtual session, floor shadowing, and manager coaching.

Frontline onboardingResponsive designBlended curriculumMicrolearning
ClientHarborline Logistics (fictional)
AudienceNew shift leads, promoted from the floor
My roleCurriculum design, module build
Seat time22 min async + 90 min vILT
Business problem

New leads were being promoted on Friday and running shifts on Monday.

Peak-season growth meant promoting 30+ associates to shift lead per quarter across four sites. Onboarding was a binder and a week of "follow Maria around." First-90-day attrition among new leads was the highest in the building, and missed carrier cutoffs spiked in their first month.

Curriculum architecture

Each modality does the job it's best at.

Program goalNew leads run an independent shift by day 30
  1. Async module (this course)Shared vocabulary, the handoff, the huddle structure. Done before week one, on any device.
  2. vILT: Running the HuddlePractice and peer feedback. The things self-paced modules can't do well.
  3. Floor shadowingTwo shifts with a certified lead, guided by an observation checklist.
  4. Manager coachingA rubric-based check-in after the first solo shift, with 30-60-90 milestones.
Design decisions

Built for a phone in a break room.

Mobile first. Most learners take it on their phone during a break, so every block reflows to a single column and works with touch.

Practice where it matters. The knowledge check and the heat-illness conversation focus on the judgment calls that cause incidents, not on trivia.

Gated, but not locked down. Continue dividers pace the content. Only the knowledge check and the floor map must be completed to move on. Everything else stays skimmable for experienced associates.

Measurement plan

What we'd track after launch

Day 30first independent shift (down from ~day 45)
−25%first-90-day attrition among new leads
−30%missed carrier cutoffs in new leads' first month

Concept project. Harborline Logistics is fictional. Facilitator guide, participant workbook, and coaching rubric are in progress.

04 · Restaurant product launch kit

Honey Hot Tenders

A complete launch kit for a limited-time menu item at a multi-unit chicken brand: a mobile training module, a bilingual build sheet laminate, an observed certification, and the Absorb LMS rollout plan.

Restaurant operationsProduct launchCertificationAbsorb LMSEnglish / Spanish
ClientGolden Hen Kitchen (fictional)
AudienceTeam members, operators, field team, corporate
My roleLearning plan, module, laminate, LMS rollout
Timeline6 weeks, recipe to launch
Business problem

An 8-week menu item has to be right in every store on day one.

Limited-time items succeed or fail in their first two weeks. Stores are run by independent operators with high-turnover, partly Spanish-speaking crews. Culinary had a recipe card. Operations needed every team member building to spec and following hold and allergen rules, with proof in the LMS before launch.

Kit architecture

Learn, reference, certify, roll out.

Launch goal95% of team members certified three days before launch
  1. Launch training moduleMobile-first course on the station, the six-step build, hold rules, allergens, and suggestive selling
  2. Bilingual build sheet laminateSingle editable source with English, Spanish, and bilingual print views
  3. Practical build checkObserved by operators and field coaches; critical standards must pass
  4. Absorb rollout planEnrollment rules, task approvals, certificate, notifications, reports, timeline, risks, handoff
Design decisions

One rule, reinforced everywhere.

"Glaze to order" protects quality and cuts waste, so the module, laminate, and practical check all reinforce it. The sorting activity specifically tests that glazing comes after holding.

The laminate does the remembering. The module explains why each step matters. Pump counts, hold limits, and allergens live on the laminate at the glaze station, where people actually use them.

Certification is observed, not quizzed. Knowledge checks gate the module, but certification comes from a short, observed build signed off in Absorb.

Measurement plan

What we'd track after launch

95%+certified by L − 3, reported by location
0critical food-safety findings in launch audits
< 5%item waste in the first two weeks

Concept project. Golden Hen Kitchen, its products, and all figures are fictional.

How I work

Start with the business problem. End with a behavior you can measure.

I partner with busy SMEs to separate what people need to do from what's nice to know, then build the smallest solution that changes the result.

01AnalyzeWhat metric is off, and is training actually the fix?
02Action mapObservable behaviors agreed with SMEs before any content gets written.
03PrototypeText storyboard and clickable prototype for early SME review.
04BuildProduction build, accessibility pass, LMS packaging and testing.
05MeasureTie results back to the metric we started with.

Fluent in Storyline and Rise. Fast in code.

I build in Articulate Storyline 360 and Rise 360 when that's your team's stack. The projects here are built web-native with an AI-accelerated workflow, using the same patterns those tools use.

The result: faster iteration with SMEs, interactions that aren't limited by a template, and logic you can inspect. Open Designer view on the scenario to see it.

Storyline variables & triggersTracked state + event logic (see Designer view)
Layers & statesFeedback overlays, hotspot states
Screen recording Show / Try / TestData-driven simulation engine
Rise blocksProcess, accordion, tabs, flashcards, labeled graphic, knowledge check, scenario
SCORM / xAPIPackaging for your LMS (next build step)
Articulate Storyline 360Rise 360Adobe Premiere ProAfter EffectsCamtasiaMicrosoft CopilotGeminiAI agentsHTML / CSS / JavaScriptTableauEnterprise LMS / LXPAction Mapping
About

Hi, I'm William.

I've spent more than 15 years where media production meets workforce learning. I started in short-form video, where a channel I ran grew to 17 million views. Since then I've led learning strategy and enablement for enterprise teams, from 40,000 frontline retail associates to C-suite leaders.

I partner closely with executives and subject matter experts to find the few behaviors that move the business, then build the learning around them. Through my consultancy, Relevant Training, I help organizations adopt generative AI tools like Microsoft Copilot, and I use an AI-accelerated workflow to build projects like the ones on this page.

I hold an Executive MBA from the University of Georgia's Terry College of Business and a degree in motion graphic design, so I think about both the business case and the craft.

Industries

Healthcare technology · Retail · Telecommunications · Financial services · Enterprise software & cloud · Sports technology · Digital media

Organizations I've worked with

QGenda · Tractor Supply Company · Cox Communications · Verizon · Salesforce · AWS · FirstTech Credit Union · Teamworks

I'm currently taking contract engagements in scenario-based compliance, systems and software adoption, blended onboarding, and AI enablement. Based in metro Atlanta and working remotely.

Available for contract work

Have a rollout, a deadline, or more projects than people?

Tell me about the project. I usually reply within one business day.