Learning That Holds Together
Permata Code was built on the belief that AI education works best when every concept is shown in clear relation to every other. We are here to make that structure visible.
Back to HomeHow Permata Code Came Together
Permata Code started in 2021 when a small group of software engineers and data practitioners in Kuala Lumpur kept noticing the same problem in the people they mentored. Learners were finishing online courses with fragments of knowledge — good fragments, but disconnected. They knew how to run a model, but not why the training loop worked the way it did. They could write Python, but could not trace how data moved through a pipeline.
The founding team spent several months mapping out what an AI curriculum would look like if the relationships between concepts were given as much attention as the concepts themselves. The result was a lattice structure — a grid where each cell holds one idea and each row connects related ideas. That structure became the backbone of Permata Code.
We opened our first cohort in Bangsar in early 2022 with twelve learners and two instructors. The cohort was small by choice. We wanted to see whether the lattice format held up under real questions, and we needed to be close enough to learners to adjust when it did not.
Since then we have run over thirty cohorts across three programme tracks. The curriculum has been revised after every few cohorts based on what learners found unclear or where they lost confidence. The structure has stayed the same; the detail within it has become sharper.
Make AI development accessible to working adults in Malaysia through structured, honest teaching.
Jalan Maarof 62, 59000 Bangsar, Kuala Lumpur
2021, Kuala Lumpur
30+ cohorts across three tracks
People Behind the Programmes
A small team of practitioners who have spent years working in data and software engineering roles across Southeast Asia.
Ahmad Zulkifli
Lead Instructor · AI Foundations
Former data engineer at a regional fintech firm. Ahmad designed the Python and data handling modules that open every track at Permata Code. He runs the weekly clinics for AI Foundations cohorts.
Nurul Rashidah
Senior Instructor · Machine Learning Practice
Eight years working with ML pipelines in logistics and supply chain analytics across KL and Singapore. Nurul leads the ML Practice track and reviews learner project submissions personally.
Kavitha Waran
Instructor · Deep Learning Framework
Kavitha spent five years in computer vision research at a university lab before joining Permata Code. She guides capstone projects in the Deep Learning Framework track and maintains the alumni space.
How We Keep the Bar Where It Belongs
Curriculum Review Cycle
Every track is reviewed after each cohort ends. If more than two learners flag the same point of confusion, that section is revised before the next cohort opens.
Cohort Size Limits
We cap enrolment so instructors can give individual attention. No cohort exceeds a size that would prevent meaningful feedback on project work.
Data Privacy
Learner information is held securely and used only for course administration. We do not pass personal data to third parties for marketing. Full details in our Privacy Policy.
Up-to-Date Material
Course materials reference current tooling and library versions. When a major framework update changes recommended practice, the relevant module is updated before the next cohort.
Clear Communication
Learners receive written feedback on project submissions. Feedback is specific to the work, not generic praise. We aim for comments that give clear direction for what to try next.
Completion Records
Learners who complete a track receive a course completion record from Permata Code. Records are issued after the final project submission has been reviewed and accepted.
Teaching AI Development in Malaysia
Permata Code operates from Bangsar, Kuala Lumpur, and focuses specifically on online AI and machine learning education for working adults in Malaysia. Our programmes cover Python programming, data handling, model training, neural network design, and deployment — the practical skills that data and software roles in Malaysian companies ask for regularly.
The technology sector in Malaysia has grown steadily, and with it the demand for professionals who can work with data and build predictive systems. We see learners coming from finance, logistics, healthcare administration, and software development backgrounds — people who already have work experience and want to add AI skills without pausing their careers.
Our three tracks are structured as a clear progression. AI Foundations gives someone with no coding background a working knowledge of Python and the language of machine learning. Machine Learning Practice builds on that foundation with applied data work and two real projects. Deep Learning Framework takes an experienced developer through the architecture and training of neural networks, finishing with a capstone project of meaningful scope.
We use a lattice framework across all three tracks — a grid structure where each concept is placed in visible relation to the ones before and after it. Learners often tell us this is the most useful thing about studying with us: they understand not just what they are doing, but where it sits in the broader picture.
Curious About a Specific Track?
Drop us a message and we will help you work out which programme fits where you are now.
Send an Enquiry