Benefits | Synapsia — AI Development School

[ pane 01 / core benefits ] reading time ~3 min

Six things that shape how the courses work

Instructor expertise

Instructors have spent years writing implementations from scratch, not just calling library functions. The Foundations course was designed by someone who built their first transformer in 2019, working through the original paper component by component.

Practical engineering background, not academic generalisation
Course material built from direct experience with the failure modes
Active practitioners, not course creators who stepped away from the work

Code-first teaching

Every concept is introduced through working code, read before it is run. The repository is the primary teaching material — the slides, where they exist, are secondary.

No black-box libraries before you understand what they wrap
Implementations grow incrementally across the course
Matrix shapes explained before any call to a framework

Written feedback

Exercise submissions and artefacts receive written responses that address the specific work submitted. Students can re-read that feedback weeks later, when a related issue comes up in a new context.

Specific to what you submitted, not a generic rubric
Reviewers read the code, not the description of the code
Five artefacts reviewed line by line in the Practitioner Programme

Conversation before enrolment

Enrolment starts with a message, not a purchase. For the Practitioner Programme, a technical conversation precedes any commitment — either side can decline to proceed.

You are not buying something before you know it is right for you
Self-check exercises published for every programme
Honest about fit, not focused on maximising sign-ups

Transparent fees

Fees are stated in Malaysian Ringgit, without upsell tiers or add-on costs. What is listed is what you pay.

Foundations: RM 1,290
Deployment short course: RM 870
Full Practitioner Programme: RM 4,650

Work stays yours

Every repository, artefact and capstone project belongs to the student. Synapsia retains no rights over what you build during a programme.

Repository transferred to your account at course end
Compute credit allowance included in Practitioner Programme
No licence restrictions on the code you wrote

→ next pane: [ 02 ] detail

[ pane 02 / detail ] reading time ~4 min

The thinking behind each benefit

[ expertise ] instructors who still work in the field

The people who run Synapsia courses have built production systems, read the literature as it was published and written implementations that had to work under real conditions. That background shapes which details get emphasised — the ones that cause problems in practice — rather than the ones that look clean in a survey. When the Deployment course covers latency budgets and where the time actually goes, that is not a theoretical concern. It is a description of what takes the most debugging time in production inference.

[ technology ] staying close to the actual code

The most durable way to understand an AI system is to read the code that implements it, not a diagram of what the code does. Synapsia courses use PyTorch, plain Python and, where relevant, the command line and Docker — tools that are readable and that expose what is actually happening. The Foundations course uses no high-level ML libraries for the first four weeks. By the time a library is introduced, you already understand what it is doing.

[ service ] small cohorts, direct access to instructors

Synapsia keeps cohort sizes small. Sessions are live, questions happen in real time and the cohort channel is monitored between sessions. In the Practitioner Programme, fortnightly one-to-one office hours are part of the structure. The school is not a platform trying to scale; it is a small operation trying to teach the material well.

[ value ] fees reflect what is included

The fees cover live sessions, recordings, exercise sets with written feedback, repositories and, in the Practitioner Programme, five artefact reviews, capstone review, office hours and a compute credit allowance. Nothing is sold separately. The full scope of what you get is listed in the syllabus before you commit to anything.

[ results ] what you take from a programme

Synapsia describes a syllabus and a workload. It does not make claims about employment, earnings or any outcome beyond the course itself. What you take from a programme is the knowledge you built and the work you produced. By the end of the Foundations course, you have a working (if small) transformer and a mental model that makes the literature readable. That is a measurable outcome — not a guarantee, but a description of what previous cohorts have produced by the final week.

→ next pane: [ 03 ] comparison

[ pane 03 / comparison ] typical providers vs synapsia approach

How the approach differs

[ feature comparison ]

Feature	Typical AI courses	Synapsia
Teaching method	Watch video, copy code, run notebook	✓ Write from scratch, read before running
Feedback on work	Auto-graded quiz or no feedback	✓ Written feedback on each submission
Prerequisite honesty	"Suitable for beginners" (often misleading)	✓ Self-check exercise published before enrolment
Workload stated	Vague ("go at your own pace")	✓ Hours per week and total weeks listed plainly
Session format	Pre-recorded, no live interaction	✓ Live sessions with recordings available
Outcome claims	Job-ready, salary increase, career change	✓ Syllabus and workload described; no employment claims
Repository ownership	Platform retains access or IP	✓ Repository transferred to student at course end
Fee transparency	Upsells, add-ons, tiered access	✓ Fixed fee covers all listed inclusions
Cohort size	Hundreds to thousands per cohort	✓ Small cohorts, direct instructor access

→ next pane: [ 04 ] differentiators

[ pane 04 / differentiators ] what makes this school unusual

Three things that are harder to find elsewhere

[ 01 ] scope: excluded pane

// what each course explicitly does not cover

Every Synapsia course publishes a "scope: excluded" section alongside the syllabus. This tells you what the course deliberately leaves out and why. Students who want to know before they start whether a particular topic is covered get a clear answer, not a vague "advanced topics are covered."

[ 02 ] prerequisites as checklists

// designed to be honestly failed

The self-check exercise for each programme is not a formality. It is written so that a developer who is not ready will work through it and notice where they get stuck. That is useful information to have before paying a fee and committing to six weeks of evenings.

[ 03 ] workload bar charts

// hours per week, stated as a chart

The Practitioner Programme syllabus includes a monospace bar chart of hours per week, broken down by activity type. This is not the most polished visual design, but it is the most honest one. You can look at it and decide whether those hours are available in your schedule before you start.

→ next pane: [ 05 ] milestones

[ pane 05 / milestones ] since 2022

Where the school is now

[ cohorts ]

12

cohorts completed since 2023

[ students ]

180+

developers through at least one programme

[ artefacts ]

260+

artefacts reviewed line by line

[ capstones ]

34

capstone projects reviewed and defended

[ recognitions ]

Featured in MDEC's 2024 directory of technical upskilling providers for Malaysia's digital sector workforce.

Selected for HRDC's list of approved training providers, making fees claimable under the HRD Corp levy for eligible Malaysian employers.

Invited to present the Foundations course structure at PyCon APAC 2024 as a case study in technical AI education.

[ course feedback, July 2025 ]

// aggregate from end-of-course surveys, June–July 2025

4.7/5 overall course rating, Foundations cohort (July 2025)

4.8/5 feedback quality rating, Practitioner Programme (June 2025)

94% of respondents said the stated workload matched their experience

What is different about studying here

Six things that shape how the courses work

The thinking behind each benefit

How the approach differs

Three things that are harder to find elsewhere

Where the school is now

See if one of the programmes is a reasonable fit