Reasoning together, building something real
Kasturi Tech was founded to make AI development accessible to learners in Malaysia — through structured thinking, not scattered tutorials.
← Back to HomeBuilt in Kuala Lumpur, for learners here
Kasturi Tech started in 2021 when a group of Malaysian developers noticed a gap between what AI tutorials offered and what working practitioners actually needed. The tutorials moved quickly and assumed a reader who already knew the underlying logic. That left most learners copying code without understanding what it did or why it worked.
The founding idea was simple: slow down and show the reasoning. Not just the answer, but the chain of thought that leads to it. A course where you could stop at any step, ask why, and get a real answer.
We opened our first cohort with eight learners and kept notes on every question that came up. Those questions shaped the curriculum that exists today — structured, clearly sequenced, and built around the moments where understanding actually forms.
Today Kasturi Tech runs three tracks, from a starter course for those new to Python through to an advanced deep learning programme. All are taught in small groups. All include personal feedback. All are held by people who still write and deploy models themselves.
Our Mission
To create a space where the logic of AI development is visible at every step — so learners in Malaysia can build fluency that holds beyond the course and into real work.
How We Work
Every programme at Kasturi Tech follows the same principle: show the reasoning rail alongside the code. Learners are not expected to copy — they are expected to understand. Sessions are structured as worked examples with space for questions. Projects are treated as conversations, not tests.
We keep cohorts small because the quality of feedback matters. When a learner submits a project, the response addresses their specific choices — what worked, what did not, and one clear direction forward.
3+
Years running
180+
Learners trained
12
Max cohort size
Those who design and deliver the courses
Farid Azmi
Lead Instructor — ML & Deep Learning
Farid has spent eight years building production ML systems for logistics and fintech companies in Malaysia. He designed the Machine Learning Track and the Deep Learning Reasoning Room, with a focus on making training logic legible to newcomers.
Nadia Razali
Curriculum Lead — Starter & Foundations
Nadia built the AI Starter Course from questions collected across three years of workshops. Her approach keeps each step visible and unhurried, so that the reasoning, not just the syntax, is what learners take away.
Suresh Pillay
Technical Mentor & Projects Lead
Suresh reviews every capstone project in the advanced track and runs the weekly clinic sessions. He brings a background in data engineering and a preference for feedback that is specific rather than general.
How we maintain the quality of each programme
Curriculum review cycle
Course materials are reviewed after every cohort. Where learners consistently ask the same question, the step is rewritten. Nothing is left because it has always been that way.
Instructor accountability
Each instructor is a working practitioner. Teaching duties are held alongside active development work so that the curriculum stays connected to what the field actually looks like today.
Data and privacy care
Learner data is used only for course administration. No personal information is shared with third parties for marketing purposes. Consent is collected before any optional tracking is used.
Feedback that is specific
Project feedback addresses the learner's actual work — not a template. Instructors are expected to reference specific lines, decisions, or choices when responding to submissions.
Cohort size limits
Enrolment closes when a cohort reaches twelve learners. This is not a marketing constraint — it is what keeps feedback meaningful and clinic sessions worth attending.
MQA-aligned outcomes
Our programme outcomes are mapped against the Malaysian Qualifications Agency framework for professional development. Completion records are issued by Kasturi Tech and can be shared with employers on request.
Clarity over complexity. Depth over speed.
Kasturi Tech sits in Kuala Lumpur's growing AI education landscape not as the largest provider but as one with a particular point of view: that understanding the reasoning behind a model matters more than knowing which function to call. This shapes every part of how the school runs — from the way curriculum notes are written to the length of time given to each topic.
The three-track structure reflects the belief that learning AI development is not a weekend activity. Each track is sized to the complexity of its subject. The Starter Course asks six weeks because Python and basic modelling ideas need time to settle before they become usable. The Machine Learning Track runs eleven weeks because building good intuitions about training and evaluation is patient work. The Deep Learning Reasoning Room uses thirteen weeks because deploying neural networks without understanding their failure modes is a habit the programme tries not to teach.
The school does not position itself as a pathway to any specific job or role. That claim is one the programme cannot control and, therefore, does not make. What the curriculum can do — and works to do — is place each learner in a stronger position to think through AI problems clearly, read documentation with confidence, and continue learning on their own.
Kasturi Tech's location in Kuala Lumpur is relevant in a practical sense. Session times are chosen to fit Malaysian working hours. Clinic sessions are held in the evenings. Instructors are contactable through the channels learners already use. The school runs on the understanding that the people attending it have jobs, families, and limited evenings — and that the curriculum should respect that.
Interested in studying with us?
Send a message through the contact form. We will help you find the right track and answer any questions before you decide.
Get in Touch