What learners say
Accounts from people who have been through the stages
These are accounts from learners who completed one or more stages. We have included some that mention things that were harder than expected, because that is more useful to read than uniformly positive summaries.
Back to Home3
years operating
6
cohorts completed
84%
stage completion rate
4.4
avg. satisfaction (out of 5)
Learner accounts
From past cohort participants
Hafizuddin Yusof
Data Analyst · Penang
"Stage 1 was harder than I thought it would be, particularly the statistics section. The workload estimate on the page said eight hours per week and I was closer to ten in weeks four and five. But the written feedback on exercises was what I needed — not just a score but actual notes on where my reasoning was off. I went on to Stage 2 feeling like I understood why things were working, not just that they were."
Stage 1 · June 2025
Siti Balqis
Software Engineer · Kuala Lumpur
"I came into Stage 2 having written production Python for three years so I skipped Stage 1. The model building work was more involved than I expected at weeks seven and eight when convolutional architectures came in alongside the project milestone. The mentor was genuinely useful — same person every week, knew my code by that point. I would have given five stars except the cohort study group scheduling didn't work well for me in the KL time zone."
Stage 2 · May 2025
Tan Kean Hwa
Operations Manager · George Town
"The self-check quiz before Stage 1 told me plainly that I was not ready to go straight to Stage 2. I took that seriously and started from Stage 1, which was the right call. The pace is measured but it does not feel slow — there is always something to apply the week before that appears again in the next week's material."
Stage 1 · July 2025
Nadia Abdul Rahman
Research Associate · Shah Alam
"Stage 3 was the most demanding thing I have done professionally since my postgraduate research. Fifteen hours a week over twenty weeks while employed is not something to underestimate, and the dropout note on the cohort page said exactly that — employment conflict was the main reason people left previous groups. I stayed. The partner brief was a real logistics company with real constraints. The architecture review in week ten was the most useful feedback session I have had on technical work."
Stage 3 · April 2025
Rajendran Vijayakumar
Systems Engineer · Johor Bahru
"I appreciated that the team told me Stage 2 started with training loops before any framework. I had seen other courses do it the other way and ended up with learners who could run someone else's code but not diagnose their own. Stage 2 at Neurokiln was different in that respect. The GPU credits being included was also a practical benefit — I would not have had access to that hardware otherwise."
Stage 2 · June 2025
Lim Chee Wai
Financial Analyst · Petaling Jaya
"Switching from finance into data and then AI development was something I had been thinking about for two years. Stage 1 was where I started and it was calibrated well for someone with spreadsheet experience — the early Python sessions felt appropriate rather than condescending. The statistics module was harder than I expected but the office-hours slot meant I could ask specific questions about specific moments in the exercises."
Stage 1 · July 2025
Case accounts
Learner journeys through the stages
Account 01 · Stages 1 and 2
Starting position
Hafizuddin worked as a data analyst processing structured survey data in Excel. He was comfortable with pivot tables and basic descriptive statistics but had no Python experience and had not worked with unstructured data or machine learning output.
Through the stages
He completed Stage 1 over six weeks at around nine to ten hours per week. The statistics module on reading model output was the part he spent the most time on. He entered Stage 2 having understood why the Stage 1 material was there, and his Stage 2 project was a classification model applied to data from his existing work context.
Outcome
Completed both stages over eight months while employed full-time. The completion record for Stage 2 includes the project description. He has since been working on model development tasks within his existing team, applying what he built in Stage 2 to a production context.
"The Stage 2 completion record listing the project specifically was what I needed to explain what I had done to my manager. It was more useful than a generic course name would have been."
— Hafizuddin Yusof, Stage 1 + 2
Account 02 · Stage 3
Starting position
Nadia completed a postgraduate research programme in computational biology, had working Python experience and had taken an open online course in deep learning the previous year. She could run published models but had not built one from a brief with external constraints.
Through Stage 3
The partner brief involved a logistics dataset with access restrictions and a cost ceiling for inference. Her team of three divided the requirement analysis, data pipeline and model selection work. The cost modelling workshop in week six was the point she identified as most directly shaping their architecture decision. Two architecture reviews followed at weeks nine and fifteen.
Outcome
Delivered the final presentation to the partner in week twenty. The completion record describes the specific brief and the delivered work. She reported approximately seventeen hours per week during the evaluation design and presentation preparation phases — above the stated median of fifteen.
"Working under real data constraints changed how I thought about model selection. In the open course I had done before, the data was already clean and complete. In Stage 3, it wasn't."
— Nadia Abdul Rahman, Stage 3
Reach us
Contact Neurokiln
Phone
+60 4 227 5836Address
Jalan Kelawei 76
10250 George Town
Pulau Pinang
Office Hours
Mon–Fri: 09:00–18:00 MYT
Sat: 10:00–14:00 MYT
Standards
Professional approach across all stages
Learner data protection
All records processed under Malaysia's Personal Data Protection Act 2010.
Verified partner briefs
Stage 3 briefs sourced from organisations with written agreement on scope and contact commitment.
Written feedback standard
All submitted exercises returned with written notes, not only scores, within five working days.
Honest workload data
Cohort workload figures from learner self-reports published before each enrolment opens.
Enquire
Want to ask about a specific stage before deciding?
Contact the team directly. We will go through the workload figures, the included materials and any questions about whether your background is a fit for the stage you are considering.