Where AI education
finds its structure.
Tensorhaus was built on the belief that machine learning is learnable — if the course is honest about what each concept requires and who it is written for.
The Tensorhaus story
Tensorhaus started in Kuala Lumpur from a straightforward observation: there were many people working in software, finance and operations who wanted to understand AI properly, but the available options were either too shallow or assumed a full-time student's schedule. Evening workshops kept the content light. Multi-year university programmes were not compatible with a career already in motion.
The founders — engineers with experience across model development and production deployment — put together a curriculum that treated the mathematics honestly without treating it as the destination. The point was always to get to working code that a practitioner understands well enough to debug, retrain and explain to colleagues.
The name comes from the rank-three tensor: the structure that gives the design its shape, with content organised across three depth levels the way a tensor holds values across three dimensions. It was a way of naming something that had a real structure rather than a loose collection of topics.
Mission
To make applied machine learning accessible to working adults in Malaysia through structured courses that are honest about prerequisites, transparent about workload and clear about what each programme produces.
Vision
A community of engineers across Southeast Asia who built their AI knowledge deliberately, with the mathematics in place, and who can read, write and evaluate machine learning work in a professional context.
Values
Clarity over hype. Honest prerequisites. Small cohorts. Individual feedback. A written record that reflects real work completed — not a badge for watching videos.
The people behind the courses
Each course is designed and reviewed by engineers who work on production AI systems, not academic lecturers adapting research slides for a classroom.
Ahmad Haris
Curriculum Lead & Senior Instructor
Ahmad spent eight years building data pipelines and classification models at a KL-based fintech before moving into education. He designed the Foundations and Deep Learning tracks and handles most of the live sessions.
Siti Farhana
Capstone Mentor & ML Engineer
Siti works as a machine learning engineer and mentors Capstone learners one-to-one each week. She brings a practical focus on deployment, evaluation and the engineering work that surrounds a trained model.
Raj Kumar
Tutor — Foundations Track
Raj reviews code submissions for the Foundations cohorts and runs the cohort forum. His background is in scientific computing and numerical methods, which makes him well suited to the mathematics sections of the course.
How we maintain course quality
Every part of the curriculum goes through a defined review process before it reaches learners. These are the standards we hold ourselves to.
Content review cycle
Curriculum materials are reviewed after every cohort. If learner questions reveal a consistent gap in an explanation, the material is revised before the next intake.
Practitioner-written exercises
All graded exercises are written and tested by working engineers. They reflect problems encountered in real data work, not constructed toy examples.
Individual code review
Every submission receives written feedback from a tutor on the specific code submitted. Automated grading is not used for assessed work.
Data and privacy handling
Learner data is held only for course administration. We do not share enrolment information with third parties or use it for purposes beyond course delivery and communication.
Transparent prerequisites
Each track states clearly what background knowledge is expected. We do not enrol learners who are unlikely to engage with the material at the level the course requires.
Written completion records
Records of course completion are issued on completion of all assessed work. They state the specific topics covered and the total hours committed — not a generic certificate.
Applied AI education for professionals in Malaysia
Tensorhaus operates from Kuala Lumpur and serves learners across Malaysia who want to develop a working knowledge of machine learning and deep learning within a schedule that fits around full-time employment. The curriculum covers the areas that come up repeatedly in production AI work: data preparation, model selection, training stability, evaluation methodology, and the engineering decisions that determine whether a trained model is actually useful in a deployment context.
The three tracks are designed to follow one another. The Foundations course builds the mathematical and coding baseline — not to the depth of a research programme, but to the depth where a practitioner can read documentation, adapt existing code and understand why a model is behaving a certain way. The Applied Deep Learning Track adds the architecture knowledge and tooling fluency that production AI work requires. The Capstone programme is where that knowledge gets applied to a real project, with a mentor involved at every stage.
Working hours are designed around Malaysia's standard week. Evening sessions run on weekday evenings and the amount of coursework per week is stated clearly in each track description. This is not self-paced work — there are cohort schedules, deadlines and live sessions — but the schedule is built to coexist with a day job rather than compete with it.
The written record issued at the end of each track is specific rather than symbolic. It names the topics covered, the assessed exercises completed and the total hours committed. For learners applying for roles in data science, ML engineering or technical product work, it provides a document that says something concrete rather than an image of a badge.
Find out which track fits your background
Send an enquiry and we will share the current cohort schedule and intake requirements for the track you are considering.
Get in Touch