MyLMMCenter: a platform for teacher training and student practice
A platform I am designing and building for a specific problem in Ghana: limited teacher access to educational technology. Built from interviews with 80 teachers and 50 students, it meets teachers’ training needs and gives students more engaging practice, with accessibility built in, analytics for continuous improvement and a planned Kirkpatrick evaluation.
- Role
- Designer and developer, working alone
- Audience
- Teachers and students in Ghana, adaptable to other learners
- Status
- In development, pilot planned by the end of 2026
- Evaluation
- Kirkpatrick evaluation planned at 3 and 12 months of rollout
MyLMMCenter is in development and is not publicly available, so no screenshots are shown. The plan below shows how I am testing and releasing it.
How it will be tested and released
Testing happens in stages, and feedback stays inside the product. Select a stage to read what happens.
1. Teachers test the teacher features
A small group of teachers tests the features built for them.
2. Learners test the learner features
Selected learners then test specific learner-facing features.
3. I improve the product
What testers report shapes the next version.
4. Gradual release
I release it behind feature flags, so each feature reaches users only when it is ready.
5. Feedback inside the product
Users can send feedback from within the platform, so improvement continues after launch.
My role
I am the only designer and developer. I own the whole cycle: needs analysis and user profiling, wireframes, code, the testing plan and the release.
The design challenge
Many teachers in Ghana have limited access to educational technology, and learners struggle in several subjects, especially French, with few engaging ways to practise. The challenge is a platform that teachers can use to meet their training needs and that students can use to practise, built for Ghana first and adaptable to other learners.
How I worked, step by step (ADDIE)
These stages follow the ADDIE design process, applied iteratively because the product is still being built.
- 1
Analyze
The project comes from my own experience as a teacher in Ghana, where teachers have limited access to educational technology. In 2019, my action research on teaching French through songs showed me learners’ difficulties in several subjects, and I freelance on resources for teachers and learners. For the needs analysis, I interviewed 80 teachers and 50 students in schools across Ghana’s Eastern Region. From the interviews I built user profiles for both groups. The profiles set the requirements for the product: how interactive it should be, which colors and layouts to use, how long a screen or lesson can be, how much multimedia to include, and which practice and assessment tools to offer.
- 2
Design
I set the boundaries of the project first, deciding what the first version would and would not do. Then I drew wireframes of the functionality for the teacher-facing and learner-facing features. The design carries the principles below from the start: WCAG 2.2 AA and AODA as the baseline for accessibility; a restrained, high-contrast color palette; dual coding and multimedia choices for each kind of content; and a gamification design that rewards practice. I prototype each feature and review it before building the rest. Virtual reality is on the roadmap as an optional later feature, and the design leaves room for it.
- 3
Develop
I am building the back end and the product from scratch, for self-hosting, using TypeScript, HTML, JavaScript and Python. Accessibility is built into the components, with semantic HTML, keyboard operation, visible focus, text alternatives, captions and sufficient contrast. I source or produce the multimedia, keep each screen free of extraneous cognitive load, and develop and test continuously. Each feature is prototyped, tried and reviewed before the next one is built.
- 4
Implement
Testing will happen in stages. A small group of teachers will test the teacher-facing features first, then selected learners will test specific learner-facing features. I will fix and improve what they report, then release the product gradually, behind feature flags, so each feature reaches users only when it is ready. Training and support for teachers will go with the release.
- 5
Evaluate
Feedback will be collected inside the product itself, so it is always available. I will combine it with usage analytics and short interviews to decide what to change, and I will check every release against the accessibility baseline. A formal Kirkpatrick evaluation is planned at 3 and 12 months of rollout, described below.
How I use AI to build the platform
I use AI as a development partner on MyLMMCenter. It speeds up building and testing, and I own the decisions and the review.
Front end design
I use AI to design the front end as an LMS-like learner experience for teachers and learners: layout, navigation, activities and feedback. I set the accessibility requirements and choose what to keep.
Back end
I use AI to build the back end, working from my wireframes, my boundaries for the project and the needs analysis.
Metrics and analytics
I use AI to design richer user metrics and analytics, so improvements follow how teachers and learners actually use the platform.
Feature flags and API versioning
I use AI to set up feature flags and API versioning, so a feature reaches a few teachers first, then learners, and changes do not break what already works.
Feature prototyping
I use AI to prototype features quickly, so users can try something real early and I can adjust before building further.
What stays with me
I decide what to build and why, from the needs of teachers and learners. I review and test what AI produces before it reaches users, and accessibility stays a requirement at every step.
Principles behind the platform
Needs analysis and action research
The 2019 study on teaching French through songs and interviews with 80 teachers and 50 students defined who the users are and what they need before the design began.
Accessibility (WCAG 2.2 AA and AODA)
Accessibility is a requirement of every component, not a final check: semantic HTML, keyboard use, visible focus, text alternatives, captions and enough contrast.
Dual coding
Key ideas are paired as words plus an image or diagram, for example a French word with a picture and audio of its pronunciation, so learners have two routes to remember it.
Color theory
A restrained palette with strong contrast. Color never carries meaning alone: it is paired with text and icons, and the palette is checked for color-vision differences. Warm accents mark actions, and calm tones mark content.
Multimedia (Mayer’s principles)
Short audio and video segments with captions and transcripts, narration kept close to the visual it explains, and no decoration. Files are kept light so they load on slow connections.
Gamification
Points, streaks, badges and levels reward regular practice and mastery, not speed. Learners see their own progress, and public rankings are not the default, so the game motivates without shaming.
Virtual reality (explored)
Optional immersive scenarios, such as a virtual market for speaking French, are on the roadmap for later. They are not in the first version, because access to headsets is limited, and every scenario will also have a non-VR version.
Analytics for continuous improvement
Usage metrics show where users succeed, struggle or drop out, so each release is improved from evidence and not guesswork.
Kirkpatrick evaluation (planned)
A formal evaluation of reaction, learning, behaviour and results is planned at 3 and 12 months of rollout.
Iterative prototyping
Building and testing continuously, so each decision is checked against real use.
Staged testing and release
Teachers first, then learners, then a gradual release behind feature flags, so a problem reaches few users and can be switched off.
Built to adapt
Designed for one problem in Ghana, but structured to serve several institutions at once (multi-tenancy) and other learners later.
Accessibility, analytics and evaluation
Three commitments run through the whole project. The platform is accessible by design, it measures how people use it so it can keep improving, and its impact will be evaluated formally after rollout.
Accessibility in every release
WCAG 2.2 AA and AODA are the baseline. Every release is checked with automated tests, keyboard-only use and screen readers, and users have a clear way to report accessibility problems from inside the product.
Analytics for continuous improvement
Usage metrics, such as how many learners complete practice, how often they return, where they leave a task and which features teachers use, guide each improvement. Combined with in-product feedback and short interviews, they show what to fix next.
Kirkpatrick evaluation at 3 months (planned)
Level 1, reaction: how teachers and students rate the platform and its usability. Level 2, learning: what they can now do, from practice and assessment results. Early signs of Level 3 are also checked.
Kirkpatrick evaluation at 12 months (planned)
Level 3, behaviour: whether teachers still use the platform in their teaching and students keep practising. Level 4, results: the effect on teaching practice and on learners’ progress, for example in French.
In a workplace setting
Launching a new learning tool inside an organization follows the same path: profile the users, prototype, pilot with a small group, release gradually behind feature flags, and keep collecting feedback.
What this project shows
- I identified a real problem from my own experience and researched it with users before building.
- I carry a product from needs analysis to wireframes, code and a staged release.
- I plan testing and feedback into the product from the start.
- I plan evaluation from the start: analytics for continuous improvement, and a Kirkpatrick evaluation at 3 and 12 months of rollout.
- I apply accessibility, color, dual coding, multimedia and gamification principles in one product, not in isolation.
- I write code, not only design documents.
Next case study
Back to the case studies
Hiring for a training role, or need a course built?
Hiring
I’m open to instructional design and training specialist roles, in French and English. Send me a note, or start with my résumé.
Work with me
I design accessible, interactive online courses in Articulate Rise, Storyline, Brightspace and H5P, and I train teachers on Brightspace, SharePoint and Snagit. Tell me who the learners are and what they need to do.