Tensorhaus
Tensorhaus workshop environment

// company

Careful study for engineers
who work with data

Tensorhaus was formed to fill a specific gap: structured, technically honest ML education for practitioners who already know how to build software.

Back to Home

// our story

How Tensorhaus Came About

Tensorhaus was set up in Kuala Lumpur by a small group of engineers who had spent several years working in data infrastructure and production ML systems. Over that period, they kept encountering the same situation: junior engineers who could write clean code and work with databases, but had no clear path to learning the operational side of machine learning in a disciplined way.

Most available material fell into two groups. One group was academic — graduate-level courses focused on research methods and novel architectures, not on the practical work of putting a model into production and keeping it there. The other group made claims about career transformation that the instructors were not comfortable repeating. Neither served someone who already had a job and wanted to study ML engineering honestly.

The school's name references a mathematical structure — a tensor is a generalisation of vectors and matrices, the fundamental data type in most ML frameworks. The suffix "haus" nods to the idea of a workshop or studio: a place where craft is practised with care. That combination — the technical object and the patient workspace — describes what we are trying to build.

Tensorhaus opened its first course in 2022 with a cohort of twelve engineers from across the Klang Valley. The course materials have been revised substantially since then based on participant feedback and changes in tooling. The three offerings on this site reflect what we have found to be genuinely useful to practitioners at different points in their ML learning.

// mission

What We Stand For

Honest Prerequisites

We describe what you need to know before starting, and we mean it. Courses are not watered down to seem more accessible — they are scoped so that a prepared learner can work through them properly.

Open Tooling

Every practical element of every course uses publicly available software. We have no financial relationship with any tool vendor and no incentive to steer participants toward a platform they do not need.

Measured Claims

We describe what a course covers and what you will practise. We do not make claims about what you will earn or what positions you will obtain. That is not in our power to promise.

Sustained Pace

The weekly format is deliberate. Good technical understanding takes time. We set a pace that allows a working engineer to keep up without abandoning their current responsibilities.

// team

Instructors

AH

Ahmad Hafiz

Lead Instructor — MLOps

Six years in data engineering and ML infrastructure. Runs the MLOps course and reviews weekly lab submissions. Previously worked on model serving systems at a regional fintech.

NR

Nurul Rashidah

Instructor — Tabular Methods

Statistician by training, data scientist by practice. Teaches the Tabular Data Methods course with particular attention to cross-validation and the interpretation of tree-based models.

KW

Kian Wei

Senior Instructor — Code Reading

Backend engineer turned ML practitioner. Hosts the Friday Code Reading Hour. Reads public ML repository code on screen, explains structure and reasoning, and takes questions in the session chat.

// standards

How We Build and Deliver Courses

Lab Verification

Every lab is tested against a clean environment before each cohort. Dependency versions are pinned and verified to avoid setup failures mid-course.

Annual Revision

Course materials are reviewed and revised at minimum once per year. Sections that no longer reflect current tooling practice are rewritten, not patched.

Data Privacy

Participant data is held only as long as needed for enrolment and course delivery. We do not share contact details with third parties. Full details in our Privacy Policy.

Feedback Loop

Participants complete a structured feedback form at the end of each course. Observations that identify genuine gaps in the material are addressed in subsequent revisions.

Written Reflections

The MLOps course includes a written reflection each week. These are reviewed by instructors and returned with specific observations rather than generic commentary.

Secure Access

Course materials are delivered through an access-controlled platform. Each participant has individual credentials. Access is tied to the enrolment period.

// context

ML Engineering Education in Malaysia

Machine learning has become a practical engineering discipline in the same way that database design and network architecture have — there is a body of operational knowledge that practitioners need, distinct from the statistical theory that underlies the models themselves. Tensorhaus addresses this operational layer for engineers already working in software and data roles in Malaysia and Southeast Asia.

The Malaysian engineering labour market is well developed in software and data infrastructure. Engineers who work with pipelines, databases, and APIs are common. What is less common is strong operational knowledge of ML systems: experiment tracking, model versioning, deployment patterns that account for the particular failure modes of learned models, monitoring strategies for data and concept drift. These are the areas Tensorhaus focuses on.

The Friday Code Reading Hour addresses a different gap. Technical reading — working carefully through someone else's code and understanding the design choices behind it — is practised consistently by senior engineers but rarely taught explicitly. The session provides a low-friction weekly structure for this habit, led by an instructor who can articulate the reasoning behind what is on the screen.

All sessions and labs are conducted in English. Content is priced for the Malaysian market and designed around the schedules of working professionals in the Klang Valley and surrounding areas.

// enrol

Study with Tensorhaus

If the courses on this site fit what you are looking for, send an enquiry. We will reply with enrolment details and answer any technical questions you have before you sign up.

Send an Enquiry