Synthara Lab
AI course module overview

Programmes

Three Programmes. Three Stated Scopes. Pricing Published.

A weekend workshop, a thirteen-week course in reinforcement learning, and a twelve-month enterprise retainer. Each one carries a reading strip with the numbers, and an honest scope paragraph with the limits.

Back to Home

Instructional Approach Across All Three Programmes

Every programme at Synthara Lab starts from the same design question: what does a person need to know before choosing this course? The answer — contact hours, self-study hours, prerequisite level, assessment format — appears on a reading strip before any descriptive text.

Within each programme, the same discipline applies to references. Where a benchmark or published finding is cited, the source is named. Where a topic is not covered, it is stated explicitly rather than implied by omission.

Assessment in the longer programmes uses implementation exercises and reproduction studies. The rationale is that reproducing a published result — or failing to reproduce it and documenting why — develops more practical understanding than answering questions about what should theoretically happen.

The enterprise retainer applies these same principles to organisational settings: a written, tailored curriculum, quarterly review, and a final capability report that the organisation can use internally.

Programme 01

Weekend Workshop: Building a First Model

12 contact hrs No self-study Beginner No assessment RM 450

A two-day workshop across one weekend, taking a small tabular dataset from raw file to a trained, evaluated, and documented model. The workshop covers data exploration, data leakage, baselines that matter, a first gradient-boosted model, honest evaluation metrics, and writing the model card. All work runs in a provided sandbox — participants need working Python and a laptop.

Total contact time is twelve hours across two days. There is no pre-reading, no homework, and no assessment. The course delivers one complete path end to end at an introductory pace.

This workshop teaches one path from a raw tabular file to a documented model. It does not cover deep learning, neural architectures, image or text data, or production deployment. Those topics appear in the longer courses.

What the two days cover

  1. 1 Data loading, initial exploration, and identifying data quality issues in the provided tabular dataset
  2. 2 Detecting and handling data leakage — the most common source of misleadingly good evaluation scores in applied ML
  3. 3 Building and evaluating baselines; understanding what a useful baseline actually measures
  4. 4 Training and evaluating a gradient-boosted model against the established baselines
  5. 5 Writing the model card: what the model does, what it does not do, and how it was evaluated
Enquire About This Workshop
Tabular data exploration and model training session
  • Suitable for participants with Python experience but no ML background
  • Sandbox environment provided — no setup required beyond Python and a laptop
  • Covers a complete, real path — not a curated toy example
  • Twelve contact hours across Saturday and Sunday
Policy gradient and actor-critic architecture diagram
  • Three implementation exercises — one per major topic cluster
  • Final reproduction study with required written account of what did not replicate
  • Benchmark limitation module with all references supplied
  • Approximately 195 hours total (65 contact + 130 self-study)

Programme 02

Reinforcement Learning Fundamentals

65 contact hrs ~130 self-study hrs Intermediate 3 exercises + reproduction study RM 3,050

A thirteen-week course covering the theoretical and applied foundations of reinforcement learning. Topics: Markov decision processes, dynamic programming, temporal-difference methods, policy gradients, actor-critic architectures, exploration strategies, and the practical difficulty of reproducing published RL results.

Weekly workload is five contact hours plus approximately ten hours of laboratory work. Assessment is by three implementation exercises distributed across the thirteen weeks and a final reproduction study, which requires a written account of what did not replicate and the reasons why.

Prerequisites: a foundations course in machine learning or equivalent applied experience, plus comfort with probability. This course does not cover deep RL architectures in depth, multi-agent methods, or offline RL. It covers the documented limitations of commonly used benchmarks, with references named.

Thirteen-week topic sequence

  1. 1 Markov decision processes and the formal problem structure
  2. 2 Dynamic programming — value iteration and policy iteration
  3. 3 Temporal-difference learning (TD, Q-learning, SARSA)
  4. 4 Policy gradients and the REINFORCE algorithm
  5. 5 Actor-critic architectures and advantage estimation
  6. 6 Exploration strategies and their trade-offs
  7. 7 Evaluation under stochasticity and benchmark limitations (with references)
Enquire About This Course

Programme 03

Enterprise Training & Advisory Retainer

12-month contract Up to 20 staff/cohort Organisational Quarterly + annual report RM 4,700 / month

A twelve-month retainer for organisations building or expanding internal AI engineering capability. Designed for teams that need structured upskilling with a curriculum matched to their actual stack, not a generic course.

Included in the retainer: four quarterly training blocks (up to twenty staff each), monthly instructor office hours, a written curriculum tailored to the organisation's stack and refreshed each quarter, code review clinics against internal repositories, a documentation and evaluation template library, and a written annual capability report.

This retainer covers teaching, engineering practice review, and process design. It carries no responsibility for the organisation's production systems and makes no assessment of any regulated obligation. Those remain with the organisation's own qualified advisers. This is documented in the retainer agreement.

What is delivered across twelve months

  1. Q1 Curriculum scoping, initial training block (up to 20 staff), first code review clinic
  2. Q2 Curriculum refresh, second training block, office hours (monthly throughout), code review clinic
  3. Q3 Third training block, template library update, code review clinic
  4. Q4 Fourth training block, final code review clinic, written annual capability report delivered
Enquire About the Retainer
Enterprise AI capability building training session
  • Written curriculum tailored to your organisation's stack — updated quarterly
  • Code review clinics against your own internal repositories
  • Monthly office hours with an instructor throughout the twelve months
  • Annual capability report — a written document for internal use
  • HRDC claimable for Malaysian organisations

Which Programme Fits?

A feature comparison to help you decide between the three programmes based on your current situation.

Feature Weekend
Workshop
RL
Fundamentals
Enterprise
Retainer
Suitable for individuals
Suitable for organisations
No prior ML knowledge required Varies
Formal assessment
Curriculum tailored to your stack
Duration 2 days 13 weeks 12 months
Price RM 450 RM 3,050 RM 4,700/mo

Best for

Weekend Workshop

Someone with Python who has not yet built an ML model and wants to complete one full path in two days without committing to a longer programme.

Best for

RL Fundamentals

Someone with an ML foundations background who wants to work through RL systematically, including a formal assessment component and a reproduction study.

Best for

Enterprise Retainer

An organisation building internal AI engineering capability that wants a tailored curriculum, rotating cohorts, and code review against its own codebase across twelve months.

Standards Applied Across All Three Programmes

PDPA Data Handling

Participant data managed under Malaysia's Personal Data Protection Act 2010.

Named References

Where a benchmark or finding is cited in course material, the source is named. No unsourced summaries.

Published Syllabi

Full syllabus available before enrolment, including assessment criteria and topic week-by-week.

Quarterly Curriculum Review

Enterprise curriculum reviewed quarterly. RL course reviewed annually against current literature.

HRDC Registered

Enterprise training programmes claimable under HRDC for Malaysian-registered organisations.

Stated Scope Limits

Every programme carries an honest scope paragraph. Scope limits are reviewed whenever course content changes.

Pricing

All fees are in Malaysian Ringgit. No hidden components — what is listed is what you pay.

Individual · Beginner

Weekend Workshop

RM 450

per participant

  • 12 contact hours
  • Sandbox environment
  • No assessment, no pre-reading
  • One complete modelling path
Enquire

Individual · Intermediate

RL Fundamentals

RM 3,050

per participant

  • 65 contact hours over 13 weeks
  • ~130 hours self-study
  • 3 implementation exercises
  • Final reproduction study
Enquire

Organisational

Enterprise Retainer

RM 4,700

per month · 12-month contract

  • 4 quarterly training blocks
  • Up to 20 staff per block
  • Tailored written curriculum
  • Annual capability report
Enquire

Not Sure Which Programme Applies?

Send a short message with your background and what you want to work on. We will give you a direct answer, including if none of the three programmes is the right fit.

Contact Synthara Lab