About the school
How Tensorloom came to exist and what it is trying to do
A small school built around the idea that technical AI education works better when it is narrow in scope, honest about prerequisites, and reviewed by people who do the work.
Back to HomeOur story
Where the school came from
Tensorloom started in Cyberjaya in 2022, at a point when the distance between what online AI courses promised and what engineers could actually do after completing them was wide and largely unaddressed. The courses that existed were either shallow introductions with no practical depth, or graduate-level programmes that required a level of prior mathematics most working engineers did not have sitting ready.
The school was put together by practitioners who had spent years building and maintaining ML systems in production — people who knew from repeated experience which Python habits save you weeks later, which mathematics you need versus which you can look up, and how long it actually takes to build something that works.
The name comes from the loom as a structure: threads crossing in a fixed grid to produce something functional. The visual design of the school reflects this — a visible 12-column grid, flat SVG diagrams, no stock photography of people at laptops. It is a school about structure, and the page should look structured.
Tensorloom is registered in Malaysia and runs all sessions online. The office is in Cyberjaya, Selangor. We are small by choice: small cohorts mean questions get answered in the session rather than in a forum thread three days later.
What we stand for
The approach in plain terms
Narrow scope on purpose
Each course covers a defined set of topics and says clearly what it does not cover. Scope is not expanded to look more impressive — it is kept narrow to make the material genuinely learnable in the time stated.
Prerequisites stated first
Every course page leads with a prerequisites panel placed above the syllabus, not buried at the bottom. We include a short self-check so you can decide whether to enrol without guessing.
Work reviewed by practitioners
Projects in the deep learning track are reviewed line by line by a practising engineer, not auto-graded. Written feedback explains why something works or does not, not just whether it passes a test.
No claims about employment or earnings
We describe a syllabus, a workload, and a review process. What a learner does with the skills afterwards is entirely their own matter. We make no claims about employment or earnings outcomes of any kind.
The people behind it
Teaching staff
All instructors are working engineers who also teach. No one on the teaching staff is a full-time course creator.
Aqil Zahari
Lead Instructor — Python & ML Fundamentals
Works in NLP and data engineering. Has been running Python training for engineers in Malaysia since 2020. Teaches the Foundations course and contributes to the deep learning track curriculum.
Nurul Kamaria
Instructor — Retrieval Systems & LLM Applications
Builds retrieval and search infrastructure professionally. Designed the RAG workshop curriculum and runs each workshop session. Keeps course content updated as the tooling in this space changes quickly.
Ravi Subramaniam
Senior Instructor — Deep Learning Track
Specialises in model training and production ML systems. Reviews all four graded projects and capstone submissions in the Applied Deep Learning Track. Has spent over a decade on training infrastructure at scale.
How we operate
Standards we hold ourselves to
Syllabus accuracy
Course syllabi are reviewed before each cohort opens and updated when the tooling or standard practice has changed enough to make existing content misleading.
Data privacy
Learner data is held only for operational and administrative purposes. We do not sell or share personal data with third parties for marketing. Full detail is in the Privacy Policy.
Code quality in teaching materials
Example code in sessions and exercises is written to the same standard we would apply to production code: typed, version-controlled, and written to still run six months from now.
Feedback that is specific
Exercise feedback is written, not just a score. For the deep learning track, project reviews are delivered as written comments on the repository, not as a rubric checkbox sheet.
Cohort size limits
We cap cohort sizes and close enrolment when those caps are reached. We do not expand cohorts to take more revenue at the cost of session quality.
Honest revision of scope
If a topic is removed from a course for a future cohort, that change is noted in the syllabus with a reason. We do not quietly reduce scope without acknowledging it.
Technical education in Malaysia
AI and machine learning education grounded in engineering practice
Tensorloom operates from Cyberjaya, the technology hub of Selangor, and teaches online so that engineers across Malaysia — and in other time zones — can attend. The three programmes span different levels of commitment: an eight-week evening course on Python for ML, a one-day workshop on building retrieval-augmented generation systems, and a six-month deep learning track designed for engineers making a sustained move into the field.
The pedagogical approach is rooted in the view that narrow, well-scoped courses taught live in small groups produce more durable learning than broad, asynchronous programmes. Every course begins with a clear statement of what the learner needs to know before starting and a week-by-week syllabus written in plain English. There is no ambiguity about what is covered and what is not — the "what this course does not cover" section is present on every course page, placed above the pricing, not below it.
Send us your question before you decide
We are happy to answer specific questions about course scope, prerequisites, or scheduling before you submit an enquiry form.
Contact Us