Machine learning course benefits
▸ benefits

What you get that most
courses don't offer

Tensorloom courses are built around the things that are typically missing: a syllabus you can read before paying, assessment on what you build rather than what you recall, and workload figures you can plan around.

Key Advantages

Six things that shape how we teach

Syllabus readable before you commit

The full topic list, module schedule, assessment format, and prerequisites are online before any payment. You decide on the basis of content, not a sales description.

Assessment on submitted work

Notebooks, lab reports, and a reproduction task take the place of timed tests. Grading criteria are shared at the start of each module, not revealed at submission time.

Honest workload estimates

Each course lists weekly contact hours and exercise hours separately. The numbers are based on what past cohorts actually needed, not a marketing minimum.

Compute included in the lab course

The Deep Learning Systems course provides a GPU-accessible sandbox for the full fourteen weeks. Participants do not arrange cloud accounts or manage billing during the course.

Fit check, plainly stated

The Python for ML Practice course requires a readiness quiz before enrolment. If the result does not match the course level, the school says so directly rather than proceeding with a sale.

Employer work reviewed against real repos

The cohort programme conducts fortnightly code review against the employer's actual repository, not a toy project. The engagement is scoped first in a workshop with engineering leadership.

// expertise

Instructors who work in the field

Tensorloom instructors have backgrounds in industry roles — not only in teaching. The Python for ML Practice course is led by an instructor who worked as a data engineer at a logistics platform. The Deep Learning Systems course instructor has a research background in convolutional architectures. This means the content covers what you encounter when building systems, not only what appears in textbook exercises.

  • Industry engineering and research backgrounds
  • Assessment rubrics drawn from real engineering review practices
  • References cited by paper — sources verifiable independently

run_log.txt

ROLE: Lead Instructor — industry data engineering background
ROLE: Deep Learning — research in CNNs and distributed training
ROLE: Cohort Supervisor — ML team management experience
STATUS: All instructors active in technical roles alongside teaching

compute_env.txt

ENV: Sandbox GPU environment (Deep Learning Systems)
SCOPE: Available for full 14-week course duration
ACCESS: No cloud account or billing required from participant
TOOLS: PyTorch, profiling stack, distributed training support

// technology

Compute where the course needs it

The Deep Learning Systems course covers distributed training across a small cluster and GPU profiling — topics that require actual hardware to work through properly. Rather than asking participants to set up and fund their own cloud environments, the course provides a sandbox for the duration. Participants work on real compute without managing the infrastructure around it.

  • GPU-accessible environment for 14 weeks
  • Distributed training exercises on real cluster
  • Profiling tools included in course environment

// service

A fit check before you pay

The Python for ML Practice course requires applicants to complete a published readiness quiz before enrolment proceeds. The quiz covers coding familiarity and basic mathematical concepts at the level the course assumes. If the result suggests the course would be a poor fit — either too slow or too demanding — the school says so, and can point to a more appropriate starting point. This step exists because a course that starts at the wrong level wastes time on both sides.

  • Quiz published — not a hidden filter
  • Plain feedback on result, including if the fit is poor
  • Redirect to alternative where applicable

readiness_check.txt

TEST: Published quiz, taken before enrolment
COVERS: Coding familiarity, array thinking, basic maths
IF_MATCH: Proceed to enrolment
IF_MISMATCH: School communicates directly, no pressure

pricing_structure.txt

COURSE_1: Python for ML Practice — RM 980 / 8 weeks
COURSE_2: Deep Learning Systems — RM 3,200 / 14 weeks
COURSE_3: Employer Cohort — RM 4,700 / 12 seats / 20 weeks
COMPUTE: Included in Deep Learning Systems price

// pricing

Pricing shown before enrolment

All course prices are listed on the Solutions page alongside the syllabus. There are no additional charges for the compute environment provided in the Deep Learning Systems course. The employer cohort price is per engagement, not per seat — the figure listed reflects a standard twelve-seat group. Engagements with different scope are quoted after the scoping workshop.

  • Prices listed publicly alongside syllabi
  • Compute included — no hidden infrastructure cost
  • Cohort pricing per engagement, clarified at scoping

How we compare

Tensorloom vs. typical online ML courses

Feature Typical ML course Tensorloom
Full syllabus published before payment
Assessment by submitted work, not multiple choice
Compute sandbox provided for GPU labs
Readiness check with plain outcome feedback
Research references cited by paper name
Grading rubric shared at module start
Employer programme uses real internal codebase
Honest weekly workload estimate (not minimum)

What sets us apart

Four things we do that most courses don't

Reproduction as assessment, not presentation

The final assessment in Deep Learning Systems asks participants to reproduce a published result — not present a demo. The task involves reading the original paper, implementing the training procedure, and documenting where results agree and where they diverge.

Employer scoping workshop before cohort starts

The employer cohort does not start from a template. A scoping workshop with engineering leadership defines the problem, the codebase scope, and the expected outputs. Teaching then follows from that, not from a fixed curriculum delivered regardless of context.

Post-completion clinics included

Two structured clinics after the employer cohort closes give the engineering team a place to raise questions that came up during production use of what was built. This is not open-ended support — it is two sessions with defined scope for issues that arise in the weeks after handover.

Documentation handover, not just code

The cohort programme ends with a documented handover — architecture notes, design decisions, and the written midpoint review. Engineers who were not in the cohort can read the documentation and understand what was built and why.

Milestones

Where we are now

3+

Years running cohorts

120+

Engineers through courses

14

Employer cohorts completed

MY

PDPA 2010 compliant

MDEC EdTech Recognition 2024

Recognised by the Malaysia Digital Economy Corporation in their 2024 EdTech directory for structured technical curriculum delivery.

KL ML Practitioners Network

Active member and periodic host of the Kuala Lumpur ML Practitioners reading group, operating since the school's founding study sessions.

PDPA-compliant data handling

Enrolment data and assessment records processed under Malaysia's Personal Data Protection Act 2010, with retention schedules and access rights documented.

See the syllabus. Compare the workload.

The course pages list topics, schedule, assessment format, and pricing before any enrolment decision. Send a message if anything there needs clarification.