I had just received US$1,000 in Claude API credits through Anthropic's Claude Startups programme. Naturally, my first thought was to find something useful to build.
The idea came from a conversation with my older brother last weekend.
He works as a lecturer at a local university. As we talked, the conversation turned to AI and the kinds of everyday work it might help with.
I asked him a simple question:
What parts of your work could AI help reduce, so you have more time to teach?
That question led to a learning portal, a late night, and a first version built in about three hours.
The Work Behind the Teaching
When we think about a lecturer's job, we usually picture the classroom: explaining a topic, answering questions, and helping students understand something difficult.
But my brother described the work that happens around those teaching hours.
He has to prepare materials, arrange lessons, create exercises, and follow up on students' progress. Each task matters. Together, they also take time.
Listening to him, I started thinking about the space between one class and the next. How does a student know what to study next? How does a lecturer keep track of who is moving forward and who needs more help?
Those were practical questions. They gave me something concrete to work on.
A Simple Conversation Became an Idea
My thought was: why not build a simple LMS, or Learning Management System, to help organise that learning process?
We could bring the lessons, exercises, and progress information into one place. Diploma students could work through material at their own pace, practise what they had learned, and check their understanding.
My brother could then have a clearer view of their progress and the areas where they might need guidance.
That was the starting point for UniKL LMS, the learning portal project I have been developing around the maritime learning context of UniKL MIMET.
The goal grew directly out of that conversation: make learning easier to follow, and give the lecturer more useful information about where to focus his time.
From 10 p.m. to 1 a.m.
I started building at around 10 p.m. that night.
By about 1 a.m., I had a first version.
For roughly three hours of work, it was encouraging to have something we could look at and discuss. I used AI to help build the software. The portal itself currently has no AI features; it focuses on lessons, exercises, and student progress.
The next day, we began adding materials for the Marine Diesel Engine course. That was an important step. Once real course material enters a system, you can start asking better questions about how it should be organised and how a student will use it.
A screen can look good with sample content. A useful learning experience needs to make sense with the actual material.
What We Started Building
Here is a look at the portal so far:
A first-draft showcase of the dashboard, lesson, and quiz flow.
The dashboard gives a student a place to see their courses, check lesson progress, and resume learning.
Inside a lesson, the material is broken into smaller learning blocks. The Marine Diesel Engine 1 example includes a topic on engine types and the ship's propulsion system, along with references to the source material.
A knowledge check follows the learning material, giving students a chance to test their understanding.
Put together, the intended flow is simple:
Know what to learn. Work through the lesson. Try the exercise. Review what still needs attention.
For the lecturer, the longer-term aim is to make progress easier to monitor and to help identify where a student needs support. We still need to test how well this works in practice.
Helping Students Take More Ownership
As we thought about the learning flow, Self-Directed Learning (SDL) became an important part of the idea.
In simple terms, SDL means students take a more active role in managing their learning. They decide when to study, work through the material, recognise what they do not yet understand, and seek help when they need it.
For a diploma student, that could mean returning to a lesson after class, revisiting an explanation, and trying an exercise before the next session.
If a concept is still unclear, the student can come to the lecturer with a more specific question. That gives the lecturer a better starting point for helping them.
This independence needs structure. Students need clear instructions, suitable learning materials, and a way to recognise their next step. They also need someone to guide them when they struggle.
That is the role I hope the portal can play between classes: providing a clear path students can follow, while keeping the lecturer's guidance central to the learning process.
Learning Towards a Clear Outcome
The other part of the vision is Outcome-Based Learning (OBL).
A useful lesson should make clear what the student is expected to understand or do by the end. The activities and exercises should then help the student work towards that outcome.
For a Marine Diesel Engine topic, for example, the aim could be to identify engine types and explain how parts of the propulsion system work together. The learning material and knowledge check should support that understanding.
This connects to the wider framework of Outcome-Based Education (OBE). The Malaysian Qualifications Agency's guidance on learning outcomes describes student achievement through the knowledge and competencies developed during study.
For me, the practical design question is: can a student see what they are working towards, and can the lecturer see enough information to guide them?
A completed lesson is useful progress information. A short quiz can reveal something about understanding. Practical competence still needs appropriate assessment and the lecturer's judgment.
Keeping that distinction clear will be important as we improve the portal.
How This Connects to KPT's Direction
Thinking about SDL and OBL also led me back to the wider direction of higher education in Malaysia.
In its announcement of the Pelan Pendidikan Tinggi Malaysia 2026–2035, KPT highlighted AI as an enabler for teaching, learning, and administration, alongside human values and responsible use of technology.
The Dasar Pendigitalan Pendidikan Tinggi 2025–2030 provides a direction for integrating digital technology into higher education. Its accompanying action plan includes reducing manual work and using data to support better decisions.
I see our small experiment as one way to explore those ideas at the level of a course and a lecturer's daily work.
Can we give students a clearer learning path? Can we help them become more independent? Can progress information help lecturers decide where their attention is needed?
That is my interpretation of how a project like this could contribute to the wider direction. It remains a personal development project, with no claim of official KPT or MQA approval.
Three Hours Was the Beginning
The portal is still a first draft. It has not been fully tested.
What we have so far is enough to explore the flow and see its potential. We have not yet established how much time it will save my brother, or how well students will learn through it.
Those are the things we need to find out next.
We plan to fine-tune the lessons, improve the interface, and test the experience properly. That includes checking the accuracy of the material, whether exercises support the intended outcomes, and whether the progress information is useful to the lecturer.
We also need to see how students actually move through the portal. A step that seems obvious to the person building it may need a better explanation for someone using it for the first time.
Our target is to have it ready for the next semester's student intake. Reaching that target will depend on the testing and improvements ahead.
I am pleased with what we managed to build in three hours. Now the work is to turn that first version into something dependable enough to support real teaching and learning.
A Thank You to Claude Startups
I also want to thank Anthropic's Claude Startups programme for approving my application and providing the US$1,000 in Claude API credits.
For a small project, API credits give me room to experiment, test ideas, and explore AI-assisted development without every early attempt adding to the same cost concern.
Beyond credits, the programme also offers access to a founder community, events, and office hours with Anthropic's Applied AI team. Those opportunities can be useful when you are working out how an idea should become a practical tool.
I am grateful for the programme credits. Membership does not imply official sponsorship or endorsement of UniKL LMS.
The credits gave me a reason to start experimenting. The conversation with my brother gave that experiment a purpose.
More Time to Teach, a Clearer Way to Learn
This reminded me of the conversation I wrote about in From Excel and Hardcopy to Intelligent Asset Management. In both cases, an idea started by listening to someone describe the work they do every day.
Sometimes that is where a useful project begins: with a simple question, a real need, and enough curiosity to try building a response.
For this portal, the hope is straightforward. If we can make the learning process easier to organise and progress easier to follow, my brother may have more time to teach and guide his students.
And his students may have a clearer way to practise, build understanding, and take more responsibility for their own learning.
That is what we will be working towards as we refine UniKL LMS for the next intake.
