Study desk with notes — participant experiences at Stardrift Labs

Participant Feedback

What People Say After Completing a Programme

Collected feedback from participants across our cohorts — their words, their experience.

Back to Home

3

Programmes offered across two levels and a year format

4.7

Average participant satisfaction rating out of 5

GMT+8

All live sessions scheduled for Malaysia time

2022

First cohort completed — programmes running since

Participant Reviews

From the People Who Studied Here

WK

Wong Kai Sheng

Backend Engineer · Kuala Lumpur

"I'd been working with APIs for years and thought I had a decent understanding of language models already. The nine-week course corrected that impression fairly quickly. The evaluation section in weeks 7–8 was particularly useful — I hadn't thought systematically about output quality before. The practice notebooks were genuinely the best part."

Working with Language Models · April 2025

NS

Nurul Syazwani

Data Analyst · Petaling Jaya

"The practitioner track covered a lot of ground. My main note would be that weeks 5–6 moved quickly through fine-tuning approaches — I had to review the notebooks a few times. That said, the cohort sessions were where I actually understood the harder concepts. Having a small group meant I could ask specific questions without feeling like I was slowing things down."

Generative AI Practitioner Track · March 2025

RV

Rajan Velayutham

Software Architect · Shah Alam

"I enrolled in the year programme with some scepticism about whether a twelve-month commitment would hold my attention. It did. Working with the same mentor throughout made a real difference — Arjun understood how my thinking on the capstone had evolved from quarter to quarter. The monthly cohort sessions were also better than I expected; talking with other working professionals in similar situations was genuinely useful."

AI Engineering Year Programme · April 2025

LM

Lim Mei Xin

Product Manager · Kuala Lumpur

"I'm not an engineer, and I was nervous about whether the entry course would assume too much. It didn't. The first three weeks built up context carefully before asking you to do any Python work. By week six I was writing prompt evaluation scripts I'd never have been able to write before. The live sessions kept me from falling behind when life got busy."

Working with Language Models · May 2025

FA

Fariz Abdullah

ML Engineer · Cyberjaya

"The RAG section of the practitioner track was what I came for, and it delivered. Clear progression from basic retrieval setup through evaluation of retrieval quality. My only feedback was that the written Milestone 3 deadline felt tight in the same week as a heavy work period — but that's more about scheduling than content. I'd recommend it to anyone in a similar technical background."

Generative AI Practitioner Track · April 2025

TC

Tan Chee Wah

Senior Developer · Bangsar

"I've done a lot of online courses over the years. Most of them are recordings with no real interaction. The cohort sessions here are what makes the difference — you're talking through the material with people at a similar stage, with an instructor who clearly knows the subject. The fee is reasonable compared to what I've paid for lesser quality elsewhere."

Working with Language Models · May 2025

Case Studies

Study Journeys in More Detail

Case Study 01 · Working with Language Models

From Ad-Hoc Prompting to Systematic Evaluation

The Starting Point

A backend engineer with three years of API work had been experimenting with language model APIs informally — copying prompt patterns from documentation and forums without a clear framework for deciding what worked and why.

What the Course Provided

The nine-week course gave a structured vocabulary for what he had been doing intuitively: tokenization limits, few-shot formatting patterns, and a repeatable evaluation framework for output quality across multiple prompt versions.

After Completion

The participant applied the evaluation framework from the course to a production prompt review at his company, identifying two patterns that were producing inconsistent outputs. His final project became internal documentation used by his team.

"The evaluation section was exactly what I needed. I'd been measuring outputs subjectively. Now I have a process."

Case Study 02 · Generative AI Practitioner Track

Building a Retrieval System for Internal Documents

The Starting Point

A data analyst at a Kuala Lumpur fintech had self-studied transformer architecture and wanted to move into applied engineering. She had read about RAG systems but hadn't built one from scratch or worked with the evaluation side.

What the Track Provided

The practitioner track's RAG module walked through chunking strategies, embedding selection, retrieval quality metrics, and evaluation under distribution shift. The three written milestones gave structured checkpoints for her capstone design thinking.

After Completion

Her final project was a working prototype of an internal document retrieval system using publicly available models. The mentor review identified two architectural decisions worth revisiting, which she addressed before presenting it internally. The project has since moved to a pilot phase at her organisation.

"The written milestones forced me to articulate my design decisions before the mentor review. That was uncomfortable, and also the most useful part."

Case Study 03 · AI Engineering Year Programme

A Year of Structured Study Alongside Full-Time Work

The Starting Point

A software architect at a Shah Alam engineering firm wanted to develop a working understanding of the full AI engineering stack — from model selection through deployment considerations — but had found self-directed study difficult to sustain alongside a demanding job.

What the Programme Provided

The year programme's quarterly structure and monthly cohort sessions provided the external rhythm he needed. His senior mentor, assigned at the start of Q1, tracked his progress across the year and reviewed each capstone draft as it developed.

After Completion

He completed a capstone project on evaluation frameworks for generative outputs in engineering documentation contexts. The project was his own original work; mentor feedback over three drafts helped him sharpen the scope. He describes the year programme as the first online study he has completed in full.

"I needed someone to stay accountable to. The monthly sessions and the mentor relationship gave me that. I wouldn't have finished without the structure."

Contact

Reach Stardrift Labs

Address

Unit 22, Jalan Imbi
55100 Kuala Lumpur

Office Hours

Mon–Fri: 9:00–18:00
Sat: 10:00–14:00

Join the Next Cohort

Ready to Start Your Own Study Arc?

Send an enquiry and we'll match you to the right programme for your background and schedule.

Send an Enquiry