The Programmes in Full Detail
Everything you need to evaluate each programme before committing: content, structure, delivery, assessment and pricing.
Back to HomeHow Sparsemind Programmes Are Structured
Each programme is built around a specific gap rather than a broad topic. The mathematics programme does not try to cover all of statistics — it covers the linear algebra, probability and optimisation that appear in model work, taught through code. The data-centric track does not cover data science broadly — it covers the specific discipline of treating datasets as engineered artefacts.
Assessment is based on submitted work that a tutor reads and reviews. Completion records describe what was assessed. The programmes are fully online with recorded sessions, so the schedule fits around employment.
Code First
Concepts taught in notebooks before notation
Assessed Work
Completion based on reviewed submissions
Recorded Sessions
Watch on your schedule, not a fixed slot
Cloud Credits
Included in longer programmes
Mathematics for Practitioners
A six-week course on the linear algebra, probability and optimisation that appear daily in model work, taught through code rather than proof. Written for engineers who can build a model but cannot yet explain why it converges. Six hours a week, entirely online with recorded sessions.
What Is Covered
Vectors, matrices and the geometric view of linear transformations — as code, then as notation
Eigenvalues, SVD and their role in dimensionality reduction and PCA
Probability distributions, Bayes' theorem and the foundations of inference
Gradient descent, loss surfaces and why convergence fails in specific situations
Worked solutions reviewed and forum questions answered by instructors
Included
- — Weekly Jupyter notebooks with worked solutions
- — Tutor-staffed discussion forum
- — One-to-one office hours by appointment
- — Written record of course completion
Programme Fee
RM 440
Data-Centric Modelling Track
A thirteen-week track that treats the dataset as the primary object: labelling schemes, annotation quality, bias auditing, augmentation, evaluation slicing and iteration loops that improve data rather than architecture. Suited to teams whose models plateau despite good code. Nine hours a week.
What Is Covered
Labelling schemes and the cost of annotation inconsistency — measured, not assumed
Bias auditing: slicing evaluation sets and finding performance gaps by subgroup
Augmentation choices and how they introduce correlation rather than remove it
Three assessed audits of real datasets with tutor review on each
Iteration loops that improve dataset quality rather than change the model architecture
Included
- — Three assessed audits of real datasets
- — Tutor review on each audit submission
- — Cohort discussion forum
- — Cloud compute credits
- — Written record of course completion
Programme Fee
RM 1,720
Sparse Cohort Capstone
An eighteen-week capstone with an intentionally light meeting schedule and heavy independent build time, designed for self-directed learners. Each participant scopes, builds, evaluates and documents one system, defending it in a written report and a recorded demonstration.
What You Do
Scope a system: define the problem, dataset, approach and evaluation criteria in week one
Build across twelve weeks with fortnightly mentoring to track scope and progress
Participate in monthly cohort reviews — share progress, give and receive structured feedback
Attend technical writing and portfolio workshops in weeks fourteen and fifteen
Submit a written report and record a demonstration in the final two weeks
Included
- — Fortnightly one-to-one mentoring calls
- — Monthly cohort review sessions
- — Technical writing and portfolio workshops
- — Cloud compute credits
- — Alumni forum access (permanent)
- — Written record of course completion
Programme Fee
RM 3,790
Which Programme Fits Your Stage?
Use this table to identify the right starting point. The programmes can also be taken in sequence.
| Feature | Mathematics | Data-Centric | Capstone |
|---|---|---|---|
| Duration | 6 weeks | 13 weeks | 18 weeks |
| Weekly time | ~6 hours | ~9 hours | Variable |
| Fee (RM) | 440 | 1,720 | 3,790 |
| Cloud credits | — | ||
| Assessed audits | — | 3 audits | Project |
| Mentoring | Office hours | Forum + review | Fortnightly 1:1 |
| Alumni forum | — | Permanent | |
| Best for | Engineers who want to understand model internals | Teams whose models plateau on data quality | Self-directed builders ready for a full project |
Shared Across All Programmes
Data Privacy
Participant data is used only for programme administration. No data sharing with third parties for commercial purposes.
Annual Content Review
Programme content is reviewed by instructors every twelve months against current practice. Materials that have drifted are rewritten.
Access Continuity
Recordings and notebooks remain accessible for the programme duration and at least six months after it concludes.
Start With an Enquiry
Tell us where you are in your work and which programme interests you. We will answer any remaining questions before you decide.
Send an Enquiry