A school that treats learners as professionals, not passengers.
← Back to HomeHow Kelip Akademi came together
Kelip Akademi started from a straightforward observation: most people working in Malaysian technology companies who wanted to learn applied machine learning had two options — expensive international certifications with little local context, or self-directed online courses with no one to ask when things went wrong.
The school was established in George Town, Penang, to offer something in between. Structured enough to give learners a clear line from where they are to where they want to be. Small enough that mentors know the work each person is doing. Flexible enough to fit around a full-time job.
The name Kelip — the Malay word for the gentle flash of fireflies — was chosen deliberately. Fireflies light their immediate surroundings rather than the whole horizon. The school aims for the same: equipping people with skills they can use in the work directly in front of them, rather than training for an abstract version of their future career.
What the school is for
Kelip Akademi exists to help working professionals in Malaysia build specific, applicable AI engineering skills through structured programmes with real mentorship and honest feedback.
The school publishes its completion rates and its median cohort sizes. Prospective learners can see what the commitment looks like in practice before they decide. The programmes are priced so that a professional on a mid-level Malaysian salary can attend without treating it as a major financial decision.
SCHOOL STATEMENT
"We do not use the word proven. We show the numbers and let applicants decide."
The team behind the curriculum
Siti Liana Razali
Lead Mentor — ML Engineering
Seven years building production ML pipelines at Malaysian fintech companies. Designed the deployment module curriculum and leads the cohort mentor team.
Ahmad Khairul Nizam
Curriculum Director
Previously a data science lead at a Penang-based semiconductor manufacturer. Responsible for module sequencing and the capstone project framework in the full pathway.
Nur Bashirah Osman
Statistics & Foundations Mentor
Statistician by training, data practitioner by work. Leads the Weekend Foundations course and advises learners on whether the full pathway is the right next step for them.
Rajan Pillai
Mentor — Cloud & Infrastructure
Specialises in model serving and cloud cost optimisation. Joins the deployment module cohorts for the infrastructure and monitoring weeks, bringing direct operations experience.
Chan Wei Lin
Learner Experience Coordinator
Manages cohort scheduling, submission logistics and the feedback pipeline. First point of contact for learners with administrative or access questions during their programme.
Faizal Haron
Mentor — Deep Learning
Research background in neural network architecture, applied now to practical curriculum content. Leads the deep learning and model evaluation weeks in the full pathway.
How the school operates
Named mentor assignment
Every cohort is assigned two named mentors before the start date. Learners know who is reviewing their work before they submit it for the first time.
Four-day feedback commitment
Written feedback on every submission within four working days. The school tracks this internally and shares the actual average with prospective learners on request.
Data handling
Learner data is held only as long as needed for the programme. No information is shared with third-party marketing services. Privacy policy is published and kept current.
Transparent completion data
Completion rates, cohort sizes and median time-to-finish are tracked for every cohort and shared with applicants who request them before enrolling.
Practical hardware notes
Each module description states what a learner's laptop needs to run the exercises. No proprietary lab environment is required. All tools used are openly available.
Pause policy
The full pathway can be paused between modules. The school documents the process clearly so learners can plan around work demands rather than hoping for flexibility later.
AI engineering as a craft, not a credential
Applied machine learning work in Malaysian industry draws on a mix of statistical understanding, software engineering discipline and operational awareness that few single training programmes cover well. Data scientists who can train a model sometimes have little experience shipping it. Engineers who can containerise a service may not understand why a model drifts in production. The gap between notebook and endpoint is where a large part of real-world AI work actually lives.
Kelip Akademi's curriculum is designed around that gap. The full pathway moves from data handling through to production monitoring in a sequence that treats each module as a stage in the same project rather than an isolated topic. The deployment module takes learners who already have training experience and focuses exclusively on the production side. The weekend foundations course addresses the statistical vocabulary that makes the rest of the work more legible.
The school's approach draws on experience from manufacturing, fintech and technology services industries in Penang and the wider peninsular region. Content examples and case studies reflect the kinds of data problems that Malaysian organisations actually face — sensor data, transaction records, demand forecasting — rather than the benchmark datasets more common in international courses.
Every module description includes an honest time estimate. Twelve hours per week for the full pathway, seven for the deployment module, six hours per weekend for foundations. These figures come from tracking actual learner experience across cohorts, not from optimistic programme design assumptions.
Have questions about the school or a specific module?
Send a message. Someone will respond within one working day with straightforward answers — no sales pitch unless that is what you ask for.
Contact the School