Tensorhaus
Data science lab

// solutions

Three courses.
One clear purpose.

The Tensorhaus catalogue addresses the operational, analytical, and reading layers of ML engineering practice, each as a distinct offering suited to different learning needs.

Back to Home

// methodology

How Tensorhaus Designs Courses

Each Tensorhaus course begins with a problem that comes up repeatedly in engineering work. For the MLOps course, the problem is this: a data engineer or software engineer who understands pipelines and version control but has not studied the specific operational challenges of ML systems — experiment tracking, model artefact management, monitoring for data and concept drift, documentation patterns for reproducibility. The course addresses that problem in ten weeks at a pace that fits a full-time schedule.

The Tabular Data course addresses a different problem: an analyst or data engineer who works with tabular business data and wants to understand the modelling methods that perform well on this type of data, studied with technical discipline rather than as a sales pitch for a particular library.

The Friday session addresses something neither course covers: the habit of reading code carefully. This is an engineering skill that is practised but rarely taught. The session provides a weekly structure for it.

MLOps Fundamentals

// course 01 · 10 weeks

MLOps Fundamentals

RM 4,580

A ten-week course for software engineers and data engineers who would like a careful introduction to the operational side of machine-learning systems. The course covers experiment tracking, model packaging, simple deployment patterns, monitoring of model behaviour in production, and the documentation habits that keep a system maintainable. Each week consists of a recorded lecture, a small lab in a containerised environment, and a written reflection. The course is delivered with publicly available tooling.

// key topics

  • Experiment tracking with MLflow — logging parameters, metrics, and artefacts
  • Model packaging with Docker and standard serialisation formats
  • Simple HTTP inference endpoints — structure and reliability patterns
  • Monitoring for data drift and model performance degradation
  • Documentation that keeps a model system maintainable over time

// weekly structure

01

Recorded lecture covering the week's topic — watch on your own schedule

02

Lab notebook in a containerised environment — apply the lecture material to a concrete problem

03

Written reflection — articulate what you understood and where uncertainty remains

Prerequisites: Python, Docker basics, familiarity with CI/CD concepts. Prior ML knowledge not required.

Enquire About This Course

// course 02 · 5 weeks

Tabular Data Methods

RM 780

A five-week short course for learners who work with tabular business data and would like to study the methods that perform well on this kind of problem with technical care. Topics include gradient-boosted trees, careful cross-validation, feature engineering for tabular data, and the reading of feature-importance results. Each week consists of a recorded lecture, a notebook, and a small homework. The course is practical and notebook-led.

// key topics

  • Gradient-boosted trees — LightGBM and XGBoost, structure and tuning
  • Cross-validation done carefully — stratification, leakage prevention, evaluation
  • Feature engineering specific to tabular business data
  • Reading feature-importance and SHAP outputs correctly
  • Practical notebook organisation for reproducible analysis

// weekly structure

01

Recorded lecture introducing the week's method with worked examples

02

Notebook — apply the method to a provided dataset, modify and experiment

03

Small homework — a variation problem that tests understanding of the key idea

Prerequisites: Python, pandas, basic statistical concepts (mean, variance, distributions). Prior ML experience helpful but not required.

Enquire About This Course
Tabular Data Methods
Friday Code Reading Hour

// session · weekly

Friday Code Reading Hour

RM 220 / month

A weekly one-hour open session in which a senior instructor reads a small piece of public machine-learning code on screen and discusses how it is structured and why. The session is unhurried and educational. Attendees can ask questions in chat. The session is provided as a quiet space for working engineers to keep their reading practice steady; it is not a course or assessment.

// what happens in a session

  • Instructor selects a short, meaningful piece of public ML code
  • Code is read aloud and annotated on screen, section by section
  • Design choices and trade-offs are discussed as they appear
  • Attendees post questions in chat; instructor addresses them during reading
  • Recording provided to subscribers within 24 hours

No prerequisites beyond reading Python. No assessment, no homework. This session is a reading practice, not a course.

Enquire About This Session

// decision guide

Choosing the Right Course

The three offerings address different needs. Use this table to match your situation to what each course provides.

Feature MLOps Fundamentals Tabular Data Methods Friday Code Reading
Duration 10 weeks 5 weeks Ongoing monthly
Price (RM) RM 4,580 RM 780 RM 220/month
Best for Engineers moving into ML ops roles Analysts adding ML methods Any practitioner building a reading habit
Lectures included Yes Yes
Lab environment Docker-based Notebooks
Instructor feedback on submissions Yes Yes
Live session Weekly Friday
Session recording provided Yes, within 24h
HRDF Claimable Yes Yes

// pricing

Course Fees

// 10 weeks

MLOps Fundamentals

RM 4,580

  • 10 recorded lectures
  • 10 containerised labs
  • Written reflections with feedback
  • Async support channel
  • HRDF claimable
Enquire

// 5 weeks · best entry point

Tabular Data Methods

RM 780

  • 5 recorded lectures
  • 5 notebook labs
  • Homework with feedback
  • Async support channel
  • HRDF claimable
Enquire

// monthly subscription

Friday Code Reading

RM 220/month

  • 4–5 live sessions per month
  • Session recordings within 24h
  • Chat Q&A during session
  • No assessment or homework
  • Cancel any time
Enquire

Prices quoted in Malaysian Ringgit (RM) and inclusive of applicable taxes. Payment via FPX, credit card, or e-wallet. HRDF claims available for eligible companies on the two full courses.

// standards

Technical Standards Across All Courses

Access Control

Course materials are accessible only to enrolled participants via individual credentials. Access is tied to the enrolment period and managed securely.

Environment Testing

Lab environments are tested on clean systems before each cohort begins. Dependency versions are pinned and verified to prevent mid-course setup failures.

Data Privacy

Participant data is held only for enrolment and course delivery purposes. We do not share contact details with third parties or use them for unrelated communications.

Annual Revision

Materials are reviewed and revised annually. Sections referencing deprecated tooling or outdated practice are rewritten, not left in place with warnings attached.

Support Response

Async support channels are checked on weekday mornings. Questions that cannot be resolved in writing are addressed in the weekly office-hours slot during the course period.

Receipts and Documentation

Official receipts are issued for all payments. HRDF claim documentation is provided on request for eligible companies enrolling participants in the full courses.

// enrol

Ready to start?

Send an enquiry mentioning which course interests you and your current technical role. We will reply with enrolment details and any prerequisite notes.

Send an Enquiry