LMS Migration Project v1
A prompt-engineered migration system built to scale — preserving design standards, learning integrity, and content fidelity across hundreds of assets.
When our team adopted an AI-first learning platform, we faced a familiar challenge: hundreds of existing SCORM-based assets that needed to move — and no scalable process to do it. Rather than approach migration manually, we engineered a four-phase prompt system that assessed, optimized, and ingested content at scale. By embedding the system directly into the platform’s AI corpus, we gave designers always-current prompts without ever leaving the tool — turning migration from a bottleneck into a governed, repeatable workflow.
of existing assets migrated
AI prompt system
strict content controls enforced at every ingestion step
Hundreds of SCORM-based assets needed to move to an AI-first platform with no scalable migration process. The risk: content drift, inconsistent design standards, and learning quality degrading at scale — one course at a time, manually.
Engineer a modular, four-phase prompt system covering every stage of migration — from QA and design assessment to AI-readiness and ingestion — then embed it inside the platform so designers could access always-current prompts without leaving their workflow.
Four-Phase Migration Stack
Assessed each asset against enterprise styles and standards — flagging inconsistencies before migration introduced them into the new corpus.
Evaluated accessibility, visual design, and learning design quality. Generated improvement recommendations aligned to current best practices.
Analyzed course structure for optimal AI corpus ingestion — ensuring content was architecturally sound for the platform’s AI to surface and retrieve accurately.
Instructed the AI-LMS in design standards. Faithfully recreated layouts, substituted appropriate analogs where 1:1 conversions didn’t exist, and enforced strict copy fidelity.
Accountability Built In
A dedicated guide established designer review requirements at every phase — approval gates, change logs, and accuracy checks built into the process, not bolted on after.
Prompts deployed as a gated course inside the LMS, living in each designer’s AI corpus. Always current — no hunting through a library, no risk of outdated local copies.
A single source of truth. Prompt updates propagated automatically to all users — no versioning confusion, no drift between the guide and what designers actually ran.
Discover → Design → Build → Refine → Pilot → Scale
Assessed the existing asset library, platform capabilities, and migration risk factors to define scope and requirements.
Architected the four-phase prompt system and governance framework — mapping each phase to a clear migration quality gate.
Developed and iterated all prompts; created the governance guide and designer course embedded directly inside the LMS.
Tested prompts against real assets; refined based on output quality, content fidelity checks, and direct designer feedback.
Ran a controlled migration of a small asset subset with full designer review and change-log tracking at each step.
Rolled out org-wide with the platform-embedded delivery model in place — prompts automatically current for all designers.
Prompt Engineering · AI Workflow Design · LMS Administration · Instructional Design · Content Strategy · Accessibility Standards · Change Management · AI Governance · Process Design · Version Control · Stakeholder Communication
