About the School
We Teach AI Development the Way Engineers Learn Best
Practical, written, technically honest. No career promises, no urgency. Just structured work that leaves you with something built.
← Back to HomeOur Story
How Rentak Silika Came Together
Rentak Silika started in Johor Bahru in 2022 as a response to a particular frustration: most AI learning resources available to working engineers in Malaysia were either too shallow to produce usable skills or too abstract to connect to the systems engineers actually build at work. The name — rentak, a Malay word for rhythm or cadence, and silika, referencing the silicon substrate of computing — reflects the school's interest in the layered, rhythmic nature of well-designed technical systems.
The founders came from software engineering and applied research backgrounds, not from pedagogy or marketing. That origin shows in the programmes: the emphasis is on building a working system rather than sitting through lectures, and the feedback process treats submissions as engineering artefacts rather than homework assignments.
The school's three current offerings — Applied Language Model Engineering, the Computer Vision Module, and the Technical Readiness Assessment — were developed over roughly eighteen months of iteration with early learners. Some of those learners came in with strong Python backgrounds and needed only the applied component. Others arrived with more general programming experience and benefited from spending time on mathematics and data handling before tackling the main programmes. The readiness assessment exists because of that second group, and because the school has no interest in taking fees from people who would be better served by free public material for a few months first.
Rentak Silika operates with a small instruction team. Each programme has one primary instructor who provides written feedback on every submission. The computer vision module includes a mid-module one-to-one review as a standard element, not an optional upgrade. The school does not use auto-graders or peer assessment as a substitute for instructor review.
Instruction Team
The People Behind the Programmes
Ahmad Zahiruddin
Lead Instructor — Language Models
Seven years in backend engineering before moving into applied NLP. Designed the language model programme's evaluation harness component from his own production experience with retrieval pipelines.
Nurul Liyana
Instructor — Computer Vision
Research background in image classification and dataset construction. Leads the computer vision module and wrote the annotation practice guidelines that are now part of the published curriculum.
Siva Chandrasekharan
Readiness Assessment Lead
Spent years running technical interviews and onboarding engineers into data teams. Designed the assessment framework with the explicit aim of producing honest, usable plans rather than sales funnels.
Standards
How We Maintain Programme Quality
Written Feedback on Every Submission
Every project submission receives a written review from the programme instructor. Feedback addresses technical decisions, not just correctness.
Published Syllabuses at Two Depths
Each programme publishes a summary view and a technical view of the syllabus. Prospective learners can read both before committing to any programme.
Compute Costs Disclosed in Advance
Where GPU access is required, the indicative rental cost is stated in the programme description. No learner encounters unexpected infrastructure costs mid-programme.
Data Privacy and Confidentiality
Learner submissions and assessment data are held only as long as needed for the programme. They are not shared with third parties or used for marketing.
Published List of What We Do Not Teach
Each programme specification includes explicit scope boundaries — stating what is out of scope and why. This reduces enrolment mismatches and helps learners find the right resource.
Curriculum Reviewed After Each Cohort
Programme content is reviewed and updated between cohorts based on learner feedback and changes in the technical landscape, particularly in rapidly evolving areas like language model tooling.
Our Approach
Technical Education for Engineers Who Build Things
The field of AI development has accumulated a significant body of online content over the past several years — tutorials, notebooks, video walkthroughs, and short courses from universities and companies alike. Much of this material is technically accurate and freely available. Rentak Silika does not compete with it on breadth or volume.
What the school offers is structured depth on a small number of specific topics, with instructor engagement at every stage. The language model engineering programme, for instance, covers retrieval pipeline design, evaluation harness construction, and the cost and latency considerations that shape real production decisions. These are not topics well served by a two-hour video. They require building something, encountering problems, and receiving feedback from someone who has made similar architectural decisions in practice.
The computer vision module follows the same logic. The most technically demanding part of the module — sourcing your own dataset and managing its annotation — is also the part that most resembles real project work. The school considers that alignment between programme structure and professional reality to be more valuable than a neatly packaged curriculum.
The readiness assessment occupies a different role. It is a diagnostic tool rather than a course. Its purpose is to tell a prospective learner what they currently know, what gaps would affect their progress in a specific programme, and what the most efficient path forward looks like — which sometimes means spending time on free public material before spending money on instruction.
Rentak Silika operates in English because the primary technical literature in AI development is English-language. The school is registered in Malaysia and based in Johor Bahru, which allows it to address the local engineering context — industry structure, regulatory environment, and the kinds of systems commonly built in the region — without pretending that international context does not exist.
Next Step
Start with the Readiness Assessment
A written plan, delivered within one week, telling you plainly what to learn first and whether any of our programmes are a reasonable fit right now.
Get in Touch