↳ Three courses · Three levels · RM 490 to RM 4,510
Courses built around what the work actually asks of you.
Prerequisites stated. Effort estimated honestly. Code reviewed by a working engineer.
← Back to HomeHow the courses are structured
All three courses run part-time so you keep your existing job. The effort ranges — not minimums, not marketing figures — are based on what students across cohorts actually report. If the week is heavy, the low end applies. If you want to go deep on an assignment, you will likely hit the high end.
Each course includes graded assignments with written feedback from a senior engineer. The assignments are designed to surface misunderstandings that a multiple-choice quiz would miss. You implement things. Someone reads the implementation. You get feedback on the specific choices you made.
The three courses form a sequence, but the sequence is not rigid. The prerequisite for each course is the capability, not the course before it. If you already have the skills that the foundations course teaches, you can start at deep learning — subject to a prerequisite check.
Shared across all three courses
- Written feedback on every graded assignment
- Prerequisites stated in the first line of each module
- Self-check quiz available before enrolment
- Effort ranges based on cohort reporting
- Compute requirements stated per module
- "What this does not cover" tray in every course
- Weekly live sessions with the course engineer
def mse_loss(y_pred, y_true):
# implement from scratch, no torch.nn
return ((y_pred - y_true) ** 2).mean()
Foundations of Machine Learning Engineering
Prerequisites: can write a loop and read a stack trace in Python
A twelve-week part-time course covering Python for numerical work, linear algebra and probability as engineers use them, gradient descent implemented from scratch, and the practical business of loading, cleaning and splitting a dataset without deceiving yourself. Written for working software developers and analysts.
What is included
- Weekly live sessions with a working ML engineer
- Seven graded assignments with written feedback
- Code review of your own small model implementation
- Malaysian dataset project
- Written assessment record on completion
- Self-check quiz before enrolment
What this does not cover
Production deployment, deep learning architectures, distributed training, model serving. Those are covered in the next two courses.
Deep Learning Systems and Model Training
Prerequisites: foundations course completed, or trains models at work already
A sixteen-week course on building, training and debugging neural networks with PyTorch: dataloaders, mixed precision, distributed training across two or more GPUs, checkpointing, and the unglamorous work of finding out why a loss curve is lying to you. Suited to learners who have completed a foundations course or who already work with models.
What is included
- GPU credits on a shared cluster
- Fourteen assignments with code review on every submission
- Two open-ended projects with public write-ups
- Weekly office hours
- Reproducibility requirement for all results
- Portfolio repository you own at the end
What this does not cover
End-to-end system design, production serving, monitoring, model risk assessment, and evaluation methodology design. Those are covered in the residency.
model = DDP(model, device_ids=[local_rank])
scaler = GradScaler() # mixed precision
with autocast(): loss = criterion(out, y)
pipeline = DataPipeline(source="my_data")
model = train_with_eval_harness(pipeline)
serve(model, monitor=True) # + postmortem
Applied AI Engineering Residency
Prerequisites: production experience with deployed systems
A twenty-four-week residency in which a small cohort builds and ships a working AI system end to end: data pipeline, evaluation harness, model, serving layer, monitoring and a written post-mortem. Aimed at engineers with production experience who want to work under supervision on a system that must actually run.
What is included
- Dedicated mentor throughout the residency
- Cluster access throughout
- Weekly design reviews
- Written evaluation methodology critiqued by a second reviewer
- Model risk, dataset provenance, and consent sessions
- Public technical talk at the close
- Detailed written assessment of your engineering practice
- Portfolio review and introductions (without promises of any kind)
What this does not cover
Research-level ML, academic publishing, and anything in a regulated field. This is an engineering residency, not a research programme.
Course comparison
Decide which level matches where you are right now.
| Feature | Foundations RM 490 |
Deep Learning RM 1,720 |
Residency RM 4,510 |
|---|---|---|---|
| Duration | 12 weeks | 16 weeks | 24 weeks |
| Effort per week | 8–10 hrs | 10–12 hrs | 15–20 hrs |
| Graded assignments | 7 | 14 | Ongoing |
| Code review | |||
| Live sessions | |||
| GPU cluster credits | — | ||
| Dedicated mentor | — | — | |
| Portfolio repository | — | ||
| Public technical talk | — | — | |
| Malaysian dataset project | — | — | |
| Runs on laptop | — | — | |
| Best for | Developers new to ML | Engineers ready for PyTorch | Engineers with production exp. |
Standards applied across all courses
The operational standards that apply regardless of which course you enrol in.
Reproducibility
Deep learning and residency results must reproduce. Seeds logged, checkpoints required, hyperparameters recorded.
Student data protection
Submission data is not used for model training or passed to third parties. Records kept securely and available to the student on request.
Specific written feedback
Feedback is written for your submission specifically. Not a template. Not a rubric checkbox.
Honesty about scope
Every course lists what it does not cover. You know before enrolling what the next step after this course would be.
Fit assessment before enrolment
We will tell you if a course is not the right fit right now. We do not encourage enrolment where prerequisites are not met.
No outcomes claims
We do not make claims about career outcomes, salary changes, or job placement. Portfolio review and introductions are offered without promises.
Take the self-check quiz, then send us a question.
The quiz is on the course page. If you have read the prerequisites and are unsure, an enquiry is the right next step.
Send an Enquiry