The story behind the work.

The situation

Researchers and technical specialists were asked to present their work to funders, partners, and the public. Their expertise was unquestioned; their audiences left without the one idea that mattered.

Who it was for

Scientists, researchers, engineers, and technical subject matter experts across disciplines.

What was getting in the way

Experts organize knowledge by structure — methods, then results, then implications. Audiences need it organized by relevance. Nobody had ever taught the translation, and 'simplify it' reads to an expert as 'be less accurate'.

What I designed

A learning program that treats translation as a professional skill: audience analysis, message hierarchy, analogy construction, visual choices, and rehearsed question handling — with the accuracy constraint respected throughout.

How we would know it worked

• Blind audience comprehension checks on the participant's main message, before and after.
• Expert-rated accuracy of the translated version, guarding against oversimplification.
• Downstream outcomes participants care about: funded proposals, accepted talks, partner engagement.
• Self-efficacy shift on handling unexpected audience questions.

What this demonstrates

• SME collaboration
• Technical communication
• Knowledge translation
• Adult learning

The situation

A conservation organization had excellent science, a beloved landscape, and interpretive materials that visitors read politely and forgot. Stewardship behavior at the site was not changing, and the education team was asked to prove the programming was worth funding.

Who it was for

Park and preserve visitors, volunteer stewards, school groups, and seasonal interpretive staff — most encountering the topic once, briefly, and outdoors.

What was getting in the way

People were not indifferent. They could not connect a landscape-scale problem to a decision that was actually theirs to make, in the place they were standing, that week.

What I designed

A layered learning experience: a short field-based orientation, place-anchored interpretive moments tied to one behavior each, a volunteer steward training track, and a staff facilitation guide so every seasonal interpreter tells the same story well.

How we would know it worked

• Observed on-site behaviors tied to the specific asks, sampled across the season.
• Visitor recall of the single core message at the exit point.
• Volunteer steward retention and hours contributed year over year.
• Seasonal staff confidence facilitating questions outside the script.

What this demonstrates

• Conservation learning
• Environmental education
• Field-based learning
• Public engagement

The situation

A fictional organization with a mature online catalogue saw completion rates fall below thirty percent. The assumption in the room was that learners lacked motivation and the content needed to be more engaging.

Who it was for

L&D leadership, course owners, and the internal subject matter experts who built the existing library.

What was getting in the way

The audit found the problem was structural, not motivational: unclear entry points, objectives that promised outcomes the assessments never measured, modules far longer than the available attention window, and accessibility barriers that quietly excluded part of the audience.

What I designed

A five-lens audit — learner journey, content structure, engagement design, assessment alignment, accessibility — producing a prioritized remediation plan sequenced by effort against expected impact.

How we would know it worked

• Completion rate by module, segmented by role and device.
• Drop-off point analysis to distinguish navigation failure from content fatigue.
• Assessment-to-objective alignment score before and after remediation.
• Accessibility conformance pass rate across the remediated catalogue.

What this demonstrates

• Learning strategy
• Consulting analysis
• Evaluation
• Accessibility review

Learning objectives

• Write learning objectives describing observable learner performance.
• Design assessment evidence before selecting content.
• Apply adult learning principles to sequencing and engagement.
• Apply core accessibility practices during design rather than after.

Design decisions

• Chose five design moves over a comprehensive ID survey, because SMEs adopt small portable structures and abandon frameworks.
• Made accessibility its own lesson positioned before the summary, signaling that it is design work rather than compliance cleanup.
• Used learner-produced artifacts as the primary activity so participants leave with real course material, not notes.
• Kept theory vocabulary minimal and always attached to a decision it changes.

Assessment strategy

Learners produce an objective and its matching evidence for a module they actually teach. Knowledge checks target the discriminations SMEs most often get wrong — recall versus application, and heading style versus heading structure.

Accessibility considerations

• The course demonstrates every practice it recommends, including heading structure and non-color feedback.
• All activities are keyboard operable and screen reader navigable.
• Plain language throughout, with terminology defined at first use.
• Reduced-motion preferences are respected across all transitions.

The situation

Onboarding was a compliance checklist delivered in week one. New hires finished it knowing the policies and still spent their first month unsure who to ask, how decisions were made, or what good work looked like.

Who it was for

New employees across roles, including fully remote hires with no informal hallway learning available.

What was getting in the way

The gap was not information. It was the unwritten operating norms that existing employees absorb over months and never document.

What I designed

A digital onboarding journey covering culture in practice, scenario activities drawn from real first-month situations, manager resources, and performance support tools that stay useful after the course ends.

How we would know it worked

• Time to first independent contribution, as judged by the hiring manager.
• New hire confidence pulse at 30 and 90 days on knowing where to get help.
• Manager-reported ramp quality compared with the previous cohort.
• Voluntary turnover within the first six months.

What this demonstrates

• Corporate L&D
• Employee experience design
• Scenario-based learning

Learning objectives

• Describe the organization's mission and operating values in the learner's own words.
• Identify the correct support channel for common first-week needs.
• Apply workplace norms to realistic situations encountered in the first month.
• Select an appropriate action when expectations are unclear or conflicting.

Design decisions

• Replaced the traditional policy tour with three operating values, chosen because they predict day-to-day behavior.
• Used a tabbed interaction for values so learners control pace and can revisit without re-watching linear content.
• Wrote distractors in the scenario that are plausible and common, not obviously wrong, so the decision has real cognitive load.
• Kept every section under five minutes to fit realistically into a first week.

Assessment strategy

Formative knowledge checks with explanatory feedback after each concept, followed by a summative item that requires recognizing a behavior rather than restating a definition. Reflection prompts capture self-identified gaps and feed the manager conversation.

Accessibility considerations

• All interactions are operable by keyboard and expose state to assistive technology.
• Color is never the only signal for correct or incorrect feedback — text and icon labels carry the meaning.
• Text content meets contrast requirements at standard body sizes.
• Motion is limited and respects the learner's reduced-motion preference.

The situation

Annual privacy and compliance training was completed by everyone and applied by few. Incidents kept coming from staff who had passed the assessment weeks earlier.

Who it was for

Clinical and administrative healthcare staff completing required training between patient responsibilities, in short interrupted sessions.

What was getting in the way

Staff could recite the rules. They could not apply them in the ambiguous, time-pressured moments where the rules actually matter — a hallway question, a shared screen, a family member asking.

What I designed

Microlearning built around patient scenarios and decision activities, with knowledge checks that test judgment in context rather than recall of policy language.

How we would know it worked

• Scenario decision accuracy, tracked by scenario type to locate specific weak spots.
• Reported privacy incidents by category over the following two quarters.
• Escalation quality — whether staff route uncertain cases correctly instead of guessing.
• Completion time distribution, to confirm the module fits real shift conditions.

What this demonstrates

• Technical communication
• Compliance training
• Human-centered learning

Learning objectives

• Recognize protected health information across written, verbal, and incidental formats.
• Apply the minimum necessary standard to access, sharing, and requests.
• Identify a near-miss and take the correct first-hour actions.
• Select a compliant response to a sympathetic but unauthorized request.

Design decisions

• Chose a scrolling structure over slides because staff complete training on mobile between tasks and need to resume easily.
• Used click-to-reveal for definitions so the page stays scannable for experienced staff while remaining complete for new hires.
• Built the central scenario around a sympathetic caller — the hardest case — rather than an obvious bad actor.
• Framed reporting as system improvement instead of blame, which is the documented predictor of near-miss reporting rates.

Assessment strategy

Embedded knowledge checks with corrective feedback, a summative quiz targeting judgment rather than recall, and a reflection that asks the learner to name a personal risk point in their own workflow.

Accessibility considerations

• Semantic headings create a navigable structure for screen reader users.
• Reveal interactions are buttons with expanded state, not click-only regions.
• Feedback states use text labels and icons in addition to color.
• Content is readable at 200% zoom without horizontal scrolling.

The situation

New customers of a complex software platform completed setup but stalled before reaching the features that drive renewal. Support volume was dominated by questions the product already answered.

Who it was for

Newly onboarded software customers — administrators and end users — learning under deadline pressure alongside their normal job.

What was getting in the way

Customers could find documentation but could not sequence it. Nothing told them what to learn first, what to postpone, or how to know they were ready to move on.

What I designed

A staged customer learning journey: guided onboarding, interactive product tutorials, knowledge checks, a searchable resource library, and a certification pathway that gives internal champions something to show.

How we would know it worked

• Time from account creation to first key workflow completed.
• Support ticket volume in the deflectable categories the academy covers.
• Certification completion by account, used as an adoption health signal.
• Feature activation rates for the capabilities taught in each stage.

What this demonstrates

• Customer education
• SaaS learning
• User adoption
• Performance support

Learning objectives

• Set up a workspace aligned to an existing team structure.
• Complete the core reporting workflow unassisted.
• Select the appropriate sharing permission for internal and external audiences.
• Diagnose the three highest-volume support issues.

Design decisions

• Organized around three customer jobs instead of the product's menu structure, so the course survives interface changes.
• Wrote the simulation steps against the top support-ticket drivers rather than the full feature set.
• Used flashcards for permissions because the distinction is a recall problem, while sharing decisions are a judgment problem — so those got a scenario.
• Ended with certification framing to give customer-side champions something to show internally.

Assessment strategy

Practice tasks with immediate feedback, then a certification assessment built from real diagnostic situations. Passing requires diagnosis, not definition recall.

Accessibility considerations

• Simulation steps are provided as readable text instructions, not image-only hotspots.
• Every interactive control has a discernible accessible name.
• Focus order follows the visual reading order throughout the player.
• Progress information is announced as text, not conveyed only by a bar.

The situation

The organization purchased AI tools for every knowledge worker. Six months later, licence usage sat well below expectation and the people using the tools most were not the people whose work would benefit most.

Who it was for

Corporate employees across functions, with mixed confidence — from daily experimenters to people quietly avoiding the tools entirely.

What was getting in the way

Employees did not lack access. They lacked a way to judge which tasks were appropriate, how to get usable output, and where the real risk boundaries sat. Uncertainty resolved itself as avoidance.

What I designed

A staged adoption pathway rather than a single course: fundamentals, prompt design, workflow redesign, responsible use, and scenario practice — each stage producing a real work artifact the learner keeps.

How we would know it worked

• Scenario decisions on task suitability and data handling as the in-course performance measure.
• Sustained weekly active use by function at 30, 60, and 90 days.
• Self-reported hours returned per week, sampled rather than surveyed universally.
• Count of workflows formally redesigned by teams after the program.

What this demonstrates

• Change management
• Technology adoption
• Adult learning
• Performance improvement

Learning objectives

• Explain in plain language what an AI assistant does when it responds.
• Construct a prompt containing role, context, task, and constraints.
• Evaluate whether a given task is appropriate for AI assistance.
• Apply data minimization before using any AI tool.

Design decisions

• Led with how the technology works, because every downstream judgment depends on that one idea.
• Chose 'four-part prompt' as a single portable structure rather than a catalog of prompt tricks that age quickly.
• Made the risk scenario about a well-meaning shortcut, since that is where real exposure happens.
• Included downloadable job aids because this is a workflow skill practiced away from the course.

Assessment strategy

Applied practice producing a real prompt, plus judgment items on task suitability. There is no vocabulary quiz — knowing the term 'token' does not predict safe or effective use.

Accessibility considerations

• Plain-language explanations without unexplained jargon.
• All activities completable with keyboard alone.
• Structure uses real headings and lists for screen reader navigation.
• No time-limited interactions.