Learner experiences at Gradient Owl
[ In[1]: reviews.load() ]

What people say after completing a cohort.

Reviews from engineers, analysts and researchers who went through one of the three Gradient Owl courses. The names are real; the work described is theirs.

[ In[2]: testimonials.render() ]

Learner reviews

RH

Rozana Hamid

Data Analyst · Kuala Lumpur

"I came in knowing Excel well and having done some SQL, but Python felt like a different world. The course took ten weeks and I actually finished all the exercises — which surprised me, because I had dropped out of two self-paced courses before. The pandas material alone was worth the price. My final notebook is something I have already used at work."

Python for Data and Modelling · July 2025

FK

Fariz Kamarudin

Software Engineer · Penang

"The debugging clinic in the neural networks course was the thing I did not know I needed. I had done the fast.ai courses before and understood the conceptual side reasonably well. But when my loss plateaued and I could not figure out why, I had no systematic approach. That clinic gave me one. The written feedback on exercises was slower than I expected — sometimes took four or five days — but it was specific."

Neural Networks and Model Training · July 2025

SL

Siew Lin Tan

Research Scientist · Cyberjaya

"I enrolled in the applied systems programme after finishing the neural networks course here. The jump in complexity is real but the mentor arrangement helped. Having someone assigned throughout who knows the material well means you are not stuck waiting for a forum answer when you hit a structural problem. My capstone was a retrieval system for internal research notes — reviewed by two practitioners and much better for it."

Applied AI Systems Programme · June 2025

AM

Azwan Malik

Backend Developer · Shah Alam

"Decent course. I came in already writing Python daily so some of the early weeks felt slow. It picks up from week four onward once the vectorised thinking section starts. The forum was active and the instructor answered questions the same day most of the time. I would have liked one more session on testing analysis code specifically, but overall a solid ten weeks."

Python for Data and Modelling · July 2025

NZ

Nabilah Zulkifli

ML Engineer · Johor Bahru

"The neural networks course covered attention mechanisms properly — not just the transformer diagram but actually working through the code, which I had not found elsewhere at this level of detail. Cloud compute being included meant I was running actual training jobs from week three without any setup overhead. The training-run debugging material alone took three sessions and that was the right call."

Neural Networks and Model Training · June 2025

JO

Jonathan Ong

Product Engineer · Petaling Jaya

"I did the applied systems programme while working full time. Eleven to fourteen hours a week is a real number — some weeks I hit it, some weeks I did not, and the recorded sessions helped a lot. The architecture review clinic in week twenty-something saved me from a design decision that would have caused real latency problems. My mentor flagged it before I had committed to it. That kind of input is not available in any async format."

Applied AI Systems Programme · July 2025

[ In[3]: case_studies.run() ]

Three learner journeys in more detail

RH

Rozana Hamid — Data Analyst, insurance sector

Python for Data and Modelling · 10 weeks

Challenge

Working with Excel for all data tasks and needing to handle a dataset too large for it. Had tried learning Python from tutorials but kept abandoning the process when real data did not behave like tutorial data.

Process

Went through the ten-week cohort focusing on the pandas and data cleaning material. The instructor's written notes on her exercises identified a misunderstanding about grouped operations that had been producing wrong numbers silently.

Outcome

Finished with a cleaned claims dataset and an analysis notebook that was adopted by her team for monthly reporting. Reduced a manual process that was taking eight hours a week to under thirty minutes.

"The specific note about my groupby mistake was the turning point. I had been making that error for months."
FK

Fariz Kamarudin — Software Engineer moving into ML

Neural Networks and Model Training · 18 weeks

Challenge

Had completed the fast.ai practical deep learning course but could not diagnose what was wrong when a model stopped improving. Lacked a framework for reading training curves and debugging training runs systematically.

Process

Enrolled in the eighteen-week neural networks cohort. The debugging clinic in week thirteen gave him a checklist-based approach to diagnosing training failures. Three mentor calls helped him work through a specific architecture decision for his final project.

Outcome

Finished with a trained text classification model for internal ticket routing. Now able to read loss curves and identify the cause of a training problem within thirty minutes rather than spending days on it.

"The debugging clinic should be in every ML curriculum. I do not know why it is not."
SL

Siew Lin Tan — Research Scientist, biotech

Applied AI Systems Programme · 30 weeks

Challenge

Strong background in model training but no experience building the infrastructure around models — retrieval, evaluation, latency management — needed to put a system in front of users reliably.

Process

Completed the thirty-week programme with an assigned mentor who had production retrieval experience. Architecture review clinic helped restructure an embedding store design that would not have scaled. Shipped three systems across the course duration.

Outcome

Capstone was a retrieval system for internal research notes. Two practitioners reviewed it and gave detailed notes on latency optimisation and consent flow. The capstone was later adapted for a small internal product.

"The architecture review clinic at week twenty-four found a structural problem I had not noticed. That is exactly what you need before you build on top of something."
[ In[4]: contact.display() ]

Get in touch

Address

Unit 7-3, Icon City, Jalan SS 8/39, 47300 Petaling Jaya

Office Hours

Mon–Fri 9:00 AM – 6:00 PM MYT
Sat 10:00 AM – 2:00 PM

[ In[5]: trust.load() ]

By the numbers

340+

Course completers

3

Years running cohorts

4.7

Average satisfaction rating

91%

Final project completion rate

MDEC partner programme

Listed digital skills provider since 2023

HRD Corp claimable

All three courses eligible for levy claims

Data privacy commitment

Learner data not shared outside their cohort

[ In[6]: enrolment.open() ]

Join the next cohort

Cohorts run on a fixed schedule with limited places. Send us a message to find out when the next one starts and whether it fits your timeline.

Get in Touch