// module: solutions
Three Tracks, One Clear Progression
From software engineering foundations to a deployed system with documentation and monitoring. Each track is a complete unit of work.
← Back to Home// module: approach
How the Programme Is Structured
The three Nullabyte tracks are designed as a vertical stack. Engineering Foundations covers the software tooling that model work sits on top of. Systems and Serving covers the infrastructure that model work runs through. The Practitioner Capstone Residency is where a learner takes a real problem and delivers a complete system.
You can enter at any level that matches your current background. The tracks do not require each other as formal prerequisites — they require the underlying knowledge each one builds on. If you already work daily with containers, version control, and reproducible environments, you may be ready to start at Track 02.
TRACK-01
Engineering Foundations
7 weeks · 8 hr/week
TRACK-02 · Most popular
Systems and Serving
15 weeks · 11 hr/week
TRACK-03
Practitioner Capstone
22 weeks · varies
Engineering Foundations for AI
A seven-week course on the software engineering that model work depends on: testing, packaging, dependency management, containers, reproducible environments, and version control for data as well as code. Written for developers moving into machine learning who want their experiments to survive contact with a second machine. Eight hours a week.
What the track covers:
- Unit and integration testing for ML pipelines
- Packaging and dependency management
- Container builds and reproducible environments
- Version control for datasets and model artefacts
- Graded exercises with tutor code review
How it progresses:
- 1Set up a reproducible environment from a specification file
- 2Write tests for an existing data transformation pipeline
- 3Package the pipeline as a container that runs identically on two machines
- 4Version-control a dataset change and reproduce a prior training run
- 5Submit the final project for tutor review and receive the completion record
$ track-01 --details
duration: 7 weeks
effort: ~8 hours/week
fee: RM 495
includes: exercises, review, cohort channel, completion record
cloud_credits: not required
$ track-02 --details
duration: 15 weeks
effort: ~11 hours/week
live_sessions: 2 per cohort cycle
fee: RM 2,260
includes: cloud credits, 3 assessed deployments, review, completion record
Systems and Serving Track
A fifteen-week track on getting models into production: inference servers, batching, quantisation, caching, monitoring, evaluation in the wild, and cost accounting per request. Suited to engineers who can train a model but have not yet run one under load. Eleven hours a week with two live sessions.
What the track covers:
- Inference server design and request batching
- Quantisation and model optimisation for serving
- Caching strategies for repeated inference
- Monitoring, alerting, and evaluation in production
- Cost accounting per request and capacity planning
- Three assessed deployments with load-testing exercises
How it progresses:
- 1Deploy a model behind an inference server and benchmark latency
- 2Apply quantisation and measure the effect on throughput and accuracy
- 3Add caching and monitoring; run a load test and interpret the results
- 4Build a cost accounting report for the deployed system
- 5Complete the third assessed deployment and receive the completion record
Practitioner Capstone Residency
A twenty-two week residency in which each learner takes a problem from a partner organisation or their own workplace and delivers a working system with documentation, tests, monitoring, and a written technical report. Weekly mentoring from a practising engineer, fortnightly architecture reviews, and workshops on writing and presenting technical work.
What the residency includes:
- Weekly one-to-one sessions with a practising engineer
- Fortnightly group architecture reviews
- Workshops on writing and presenting technical work
- Cloud credits and a hosted demonstration
- Alumni forum access on completion
- Written record of course completion and technical report
How it progresses:
- 1Define the problem and agree scope with the mentor
- 2Build a minimal system and present the architecture for review
- 3Add testing, monitoring, and documentation iteratively
- 4Write the technical report and prepare the hosted demonstration
- 5Present to the cohort and receive the completion record
$ track-03 --details
duration: 22 weeks
effort: varies by project
mentoring: weekly 1:1 + fortnightly arch review
fee: RM 4,160
includes: cloud credits, hosted demo, alumni forum, completion record
// module: compare
Which Track Suits You?
Use this to identify your starting point. If you're unsure, write to us and describe your background.
| Feature / Inclusion | Track 01 | Track 02 | Track 03 |
|---|---|---|---|
| Best for | Developers new to ML infrastructure | Engineers who can train but not yet serve | Practitioners ready for a full-system build |
| Duration | 7 weeks | 15 weeks | 22 weeks |
| Fee | RM 495 | RM 2,260 | RM 4,160 |
| Cloud credits included | |||
| Live sessions | 2 sessions | Weekly 1:1 | |
| Architecture review | Fortnightly | ||
| Hosted demonstration | |||
| Alumni forum access | |||
| Completion record |
// module: standards
Standards Shared Across All Tracks
Data Privacy
Submission data is used only for assessment. It is not shared, sold, or used for purposes outside the programme. See the Privacy Policy for full details.
Assessment Criteria Published
The criteria against which each exercise is assessed are published before the exercise is due. Learners know what they are being reviewed against.
Cohort Support Channel
Every track includes access to a channel shared with the tutor and cohort. Questions are answered by the tutor, not by community members with varying experience.
Curriculum Changelog
When materials are updated between cohort cycles, a changelog entry records what changed and why. Learners from prior cohorts can see what has evolved.
Clear Enrolment Terms
Fee, schedule, assessment conditions, and completion criteria are all stated before enrolment is confirmed. See the Terms & Conditions page for the full text.
End-of-Track Feedback
Each cohort submits structured feedback at the end of the track. The results are reviewed before the next cohort is planned and influence material updates.
// module: pricing
Fees at a Glance
TRACK-01
RM 495
per enrolment
- 7-week cohort
- Graded exercises
- Tutor code review
- Cohort channel
- Completion record
TRACK-02
RM 2,260
per enrolment
- 15-week cohort
- Cloud credits included
- 3 assessed deployments
- 2 live sessions
- Tutor review & completion record
TRACK-03
RM 4,160
per enrolment
- 22-week residency
- Weekly mentor sessions
- Fortnightly arch reviews
- Cloud credits + hosted demo
- Alumni forum + completion record
// module: cta
Not Sure Which Track to Start With?
Write to us with a brief description of your current background. We will suggest a starting point with no obligation to proceed.
Send a Message