AT Akar Tensor
AI and machine learning programmes
[01]

// Programme details

Three programmes. Clear scope. Stated prerequisites.

Each programme page below follows the same order: prerequisites, weekly structure, assessment, time commitment, tooling, what is not covered, price. This makes comparison straightforward.

← Back to Home
[02]

How the programmes are delivered

All three programmes are delivered in person at the Menara Sentral premises. Live sessions are two per week for the ten-week and twelve-week programmes, with the short course running two evenings per week across three weeks. Sessions run from 7:30 to 9:00 PM on Monday and Wednesday evenings.

Each session is recorded. Recordings are accessible to enrolled participants for the duration of the programme via a private link. Access expires when the cohort ends.

Assessed work is reviewed by the instructor who taught the relevant content. Feedback is written and specific, returned within five working days of each submission deadline.

[03]

Foundations of Machine Learning

Foundations of Machine Learning

# Programme metadata

Duration: 10 weeks

Sessions: 2 × 90 min per week

Schedule: Mon & Wed, 7:30–9:00 PM

Cohort cap: 24 participants

Indep. work: ~6 hrs/week

Fee: RM 1,450

Extras: None

A ten-week evening programme covering linear algebra and probability refreshers, data handling with pandas, supervised learning from linear models through gradient boosting, model evaluation, cross-validation, and the common failure modes of leakage and drift. Does not cover deep learning or deployment.

Prerequisites

  • Comfort writing Python functions, loops, and list comprehensions
  • Ability to set up and activate a virtual environment
  • Familiarity with reading tabular data (CSV, spreadsheets)
  • Exposure to any introductory statistics material (secondary level adequate)

Assessment

  • Four graded assignments, one every two to three weeks
  • One final project reviewed line by line by the instructor
  • No timed examinations

Tooling

Python 3.10+, pandas, scikit-learn, matplotlib, Jupyter. All free and open-source. No paid API accounts required.

What this programme does not cover

Deep learning, neural networks, model deployment, cloud infrastructure, time series analysis, natural language processing, computer vision, reinforcement learning.

Enquire about this programme
[04]

Applied Language Model Engineering

Applied Language Model Engineering

# Programme metadata

Duration: 12 weeks

Sessions: 2 × 90 min per week

Schedule: Mon & Wed, 7:30–9:00 PM

Cohort cap: Not specified

Indep. work: ~7 hrs/week

Fee: RM 2,900

Extras: ~RM 120 API costs (student paid)

A twelve-week programme for developers who are building on top of existing language models rather than training them. Covers prompt structure and evaluation, retrieval-augmented generation with vector stores, tool calling and structured output, context management, cost and latency budgeting, offline evaluation harnesses, and failure analysis with real logs. Includes a running project taken from prototype to a deployed internal service.

Prerequisites

  • Working Python — able to write and debug scripts independently
  • Basic familiarity with HTTP — understands request/response, headers, JSON
  • Can use Git for version control at basic level (commit, branch, push)

Assessment

  • Running project developed across the twelve weeks
  • Milestones reviewed at weeks 4, 8, and 12
  • Final deliverable: working deployed internal service

Tooling

Python 3.10+, LangChain or equivalent, at least one commercial LLM provider API (OpenAI, Anthropic, or Google — student chooses), a vector store (Chroma or Pinecone), and a deployment target. Third-party API costs approximately RM 120 over the term, paid by the student to the provider.

What this programme does not cover

Training or fine-tuning language models, model architecture, inference infrastructure at scale, multimodal inputs, mobile deployment, iOS/Android integration.

Enquire about this programme
[05]

Short Course: Data Preparation for Machine Learning

Data Preparation for Machine Learning

# Programme metadata

Duration: 3 weeks

Sessions: 2 evenings per week (4 workshops)

Schedule: Mon & Wed, 7:30–9:00 PM

Format: 4 live workshops + recorded refs

Assessment: 2 graded exercises

Fee: RM 620

Extras: None

A three-week short course, two evenings a week, focused on the part of ML work that consumes most of the time in practice: schema design, joining messy sources, handling missing values honestly, feature construction, leakage checks, dataset versioning and documentation of provenance. Delivered as four live workshops with recorded reference material and two graded exercises.

Prerequisites

  • Basic Python — enough to load a CSV and iterate over rows
  • Basic SQL — SELECT, JOIN, WHERE, GROUP BY

Who the short course suits

Analysts who work with data daily and want to formalise their approach to pipeline quality. Also used as preparation for the ten-week ML Foundations programme by participants who want more time with the data handling material before entering the full cohort.

Tooling

Python, pandas, DuckDB, Great Expectations (basic), DVC. All free and open-source. Sessions cover setup, so prior installation is not required.

What this course does not cover

Model training, evaluation, or deployment. Streaming data pipelines. Cloud data warehouses. Spark or distributed compute. Those topics belong in the ten-week programme or subsequent material.

Enquire about this course
[06]

Choosing between programmes

Each programme is self-contained. Use this table to find the right starting point.

Criterion Data Preparation (3 wk) ML Foundations (10 wk) LM Engineering (12 wk)
Python requirement Basic (load CSV, iterate) Solid (functions, envs) Working (scripts, debug)
SQL requirement Basic (SELECT, JOIN) Not required Not required
Prior ML knowledge needed None None None (but useful)
Weekly time commitment ~5 hrs (3 weeks) ~9 hrs (10 weeks) ~10 hrs (12 weeks)
Covers model training No Yes No (uses existing models)
Covers LLM application building No No Yes
Price RM 620 RM 1,450 RM 2,900 + ~RM 120 API

// Best for

Data Preparation

Analysts who handle data daily and want to formalise their pipeline approach, or anyone preparing to enter the ML Foundations cohort.

// Best for

ML Foundations

Developers and analysts who want to understand machine learning from the inside: how models work, how to evaluate them, and where they fail.

// Best for

LM Engineering

Developers who are already building applications and want to build reliable services on top of language models, with proper evaluation and cost management.

[07]

Fees

All fees in Malaysian Ringgit. No instalment plans currently available.

Short Course

Data Preparation for ML

RM 620

  • 4 live workshops
  • Recorded reference sessions
  • 2 graded exercises with feedback
  • No third-party costs
Enquire

Ten-week programme

ML Foundations

RM 1,450

  • 20 live sessions (90 min each)
  • 4 graded assignments
  • 1 final project with line-by-line review
  • Cohort capped at 24
  • No third-party costs
Enquire

Twelve-week programme

LM Engineering

RM 2,900

+ ~RM 120 API costs (student paid)

  • 24 live sessions (90 min each)
  • Running project across full term
  • 3 milestone reviews + final review
  • Deployed service as final deliverable
Enquire
[08]

Not sure which programme is right?

Send a brief note about your technical background and what you are trying to do. We will suggest the appropriate starting point or tell you if you need more Python practice first.

Get in Touch