Case Study / Digital Learning / AWS
Converting instructor-led AWS security training into scalable async learning without losing the applied decision-making the live courses were built to support.
This initiative focused on converting four instructor-led AWS Security courses into scalable async experiences while protecting applied capability. The objective was not content migration, but performance preservation in the absence of live instructor mediation.
In live sessions, instructors close comprehension gaps in real time. In async environments, the learning design has to assume that responsibility.
AWS needed four live, instructor-led security courses converted into scalable asynchronous Digital Classroom experiences in AWS Skill Builder.
The original courses relied on instructor mediation, live demos, and real-time clarification. The objective was not only to increase access and scalability, but to maintain practical security decision-making in an environment without a facilitator.
Courses included:
Security training carries real operational consequences. Misapplied configuration decisions or misunderstood governance models can introduce production risk. The async format therefore needed to preserve applied reasoning, not just conceptual familiarity.
I led the async conversion design for all four courses. While overall course flow and lab infrastructure were inherited from the original VILT sessions and lab engineering teams, I owned:
Across the four-course series, I established a consistent async structure pattern to reduce cognitive friction and help learners move between courses more easily. Lab environments and core technical source material were developed by subject matter experts and the lab engineering team. That structure, where I owned instructional design decisions while relying on SME expertise for technical content, is close to the faculty-partnership model common in academic instructional design.
Converting technical security training from VILT to async introduced clear design risks:
Security misconfiguration has real operational impact. A direct lift-and-shift of source content would have weakened the applied nature of the training, so each conversion choice was treated as a design decision rather than a formatting step.
Although the brief was framed as a conversion request, the work still required clear instructional criteria. In practice, I treated the redesign around four design targets: preserve applied decision-making, reduce passive consumption, support self-paced navigation and review, and maintain accessibility and learner support in the absence of live instruction.
Those targets shaped sequencing, interaction choice, assessment placement, and media treatment throughout the four-course series. The result was not a direct format translation, but a deliberate async redesign aimed at protecting performance under new delivery constraints.
Challenge
Long live sessions are difficult to revisit and cognitively heavy in an async format.
Decision + Intended Effect
I broke modules into focused 10-15 minute lessons, each with a clear objective and takeaway. That improved navigability, reduced fatigue, and better supported both continuous and return-and-resume study behavior.
Challenge
Full demonstration videos duplicated guided lab support and encouraged passive observation.
Decision + Intended Effect
I removed segments that repeated lab instructions and prioritized learner execution in sandbox environments. The intended effect was to reinforce hands-on practice as the primary learning mechanism instead of narrated observation.
Challenge
Instructor-led slide narration does not translate effectively to self-paced engagement.
Decision + Intended Effect
I replaced passive lecture blocks with structured interactions that asked learners to process and apply concepts as they moved through the lesson. That increased active participation and improved conceptual processing before learners entered labs.
Challenge
End-of-module checks identify misunderstandings too late in the learning sequence.
Decision + Intended Effect
I embedded knowledge checks at decision points and paired them with feedback targeted to specific misconceptions. This surfaced and corrected errors before learners transitioned into applied lab activities.
Accessibility was treated as a core design constraint, not a post-production QA step.
All content aligned with enterprise accessibility expectations grounded in WCAG principles. This included accurate captions and transcripts for video content, meaningful alt text, readable structure, accessible contrast, and hierarchy patterns that supported screen reader navigation.
My participation in AWS's internal accessibility advocacy group (CDa11y) informed this approach. In practice, I paired self-review with peer feedback to ensure design quality and inclusivity held up in the absence of instructor mediation.
Selected screens from the converted courses. Click any screenshot to view details.
At launch, available measurement was limited to completion and learner satisfaction.
As of February 2026, the four courses hold learner ratings between 4.4 and 4.6 stars, with over 300 combined reviews across the series. Both the async versions and the original ILT offerings remain active, supporting flexible access across modalities.
While satisfaction does not measure applied security transfer, it does suggest the redesign did not create a noticeable drop in learner experience during the modality shift.
Stronger validation in future efforts would include behavioral signals such as:
Defining observable performance indicators earlier in the conversion lifecycle would allow modality decisions to be evaluated beyond perception metrics.
This initiative reinforced that ILT-to-async conversion is a design challenge, not a formatting workflow. When instructor mediation is removed, content structure, feedback timing, interaction design, and accessibility quality become the mechanism of instruction.
This work reflects how I approach technical curriculum design: protect applied performance, improve accessibility and scalability within real constraints, and make design choices based on what learners need to do, not just what content exists.
If you're hiring, collaborating, or just curious, feel free to reach out.