Master of Science in Machine Learning and Foundation Models
A research intensive degree for the small number of people who will build, adapt and evaluate frontier-class models on and for the continent. It exists because Africa is acquiring accelerated computing faster than it is acquiring the people...
The four facts
What you actually need to compare
- Award and entry
- MSc
- A
- Duration and mode
- 18 months (3 semesters)
- Hybrid. Includes 12 months of paid, assessed cooperative education.
- Cost
- UGX 3,300,000 per year
- Tier 1 laboratory classification. Includes UGX 300,000 annual laboratory premium.
- Outcome targeted
- Research scientist, research engineer, foundation model specialist
- Reviewed by 2031. If the thesis stops holding, the programme is reviewed.
Pricing Model 5.1 · The Forward Standard
Master of Science in Machine Learning and Foundation Models fee
UGX 3,300,000 per year
- Laboratory tier
- Tier 1
- Annual premium
- UGX 300,000
- Per semester
- UGX 1,650,000
- Whole programme
- UGX 4,950,000
- Approximate USD
- ≈ USD 1,303
What this fee includes
- Teaching, practice tutoring, assessment and competency verification
- The Amagezi platform and AI tutor around the clock
- All learning materials
- Every examination, including first resits
- Identity card, library, guild, sports and basic wellbeing
- Co-op placement matching and the Capability Passport
- Graduation and the statutory NCHE fee
- Online application
Employment thesis
What this programme is designed to produce
Forward publishes the destination of every programme before you enrol. If the thesis stops holding, the programme is reviewed and, where necessary, retired. Next review: 2031.
Roles
- Research scientist
- research engineer
- foundation model specialist
- evaluation and alignment researcher
- doctoral candidate
- technical lead in an applied AI laboratory.
Employer types
- Research laboratories and institutes
- frontier and applied AI organisations
- universities
- national AI capability
- and deep technology ventures.
Target outcomes we hold ourselves to
- Positioned at the top of the regional technical market with explicit international mobility, and as the principal doctoral feeder for the School.
Compulsory · The Forward Core
How the Core works →Every Forward programme, including this one, carries the same five cross-cutting requirements. They are graded, not optional.
- AI Fluency
- Quantitative Reasoning
- Ethics and African Context
- Human-Advantage Skills
- Enterprise
Includes 12 months paid work-integrated learning
Am I eligible?
Check yourself against the published requirements
7 questions written specifically for Master of Science in Machine Learning and Foundation Models. Every answer is checked against a requirement Forward has published, and the result tells you which page that requirement comes from. Indicative only, the admissions office decides on your full file.
What is the highest qualification you hold?
Curriculum structure
The same degree, in both pathways
Every Forward programme is published in both pathways: Path One, the term plan of two six-month terms, and Path Two, the block sequence of twelve four-week blocks. Same competencies, same assessments, same award. Toggle to see this programme in each.
Path One term plan shown
Path One · The Term Model
| Code | Course | CU | Hours |
|---|---|---|---|
| FDC 501 | Applied Intelligence for Professionals Both | 4 | 2-2-0 |
| FDC 502 | Research Evidence and Method Both | 4 | 2-2-0 |
| AIC 510 | Advanced Machine Learning Theory Both | 6 | 4-2-2 |
| AIC 511 | Deep Learning Architectures and Training at Scale Performance | 6 | 3-2-4 |
| AIC 512 | Research Methods, Reproducibility and Scientific Writing Both | 5 | 3-2-2 |
| AIC 513 | Data Governance and Corpus Construction Both | 5 | 3-2-2 |
| Code | Course | CU | Hours |
|---|---|---|---|
| FDC 503 | Governance, Ethics and Leadership of Intelligent Systems Both | 4 | 2-2-0 |
| AIC 520 | Foundation Models: Pre-training, Adaptation and Alignment Performance | 7 | 4-2-4 |
| AIC 521 | Evaluation Science and Both | 5 | 3-2-2 |
| AIC 522 | Efficient and Constrained Compute Modelling Performance | 5 | 2-2-4 |
| AIC 523 | Multilingual and Low Resource Modelling Performance | 5 | 2-2-4 |
| AIC 524 | Advanced Elective (safety, multimodality, RL, or applied domain) Objective | 4 | 3-2-0 |
| Code | Course | CU | Hours |
|---|---|---|---|
| AIC 530 | Research Dissertation and Model Artefact Performance | 20 | 0-2-38 |
| AIC 531 | Doctoral Preparation and Publication Seminar Objective | 5 | 4-2-0 |
| AIC 532 | Research Translation and Commercialisation Objective | 5 | 4-2-0 |
Total credit units: 90
Curriculum source
Forward University Curriculum Compendium
Assessment
Coursework + project + integrated exam
Work-integrated learning
12 months paid co-operative education
Your week at Forward
A typical week on Master of Science in Machine Learning and Foundation Models
Path One, the Term Model: your own pace across a six-month term, one live session a week, short flexible lessons, and assessment on demand.
Monday
Tuesday
Wednesday
Thursday
Friday
Curriculum explorer
Drill into the degree, step by step
Choose a study path, then a year, then a term or block, then a course. Credit totals add up as you go, and every course shows what comes before it and what it unlocks.
Step 1, choose a year
Step 2, choose a term
TERM 1 (SIX MONTHS) | FORMERLY SEMESTER 1
Year 1 · this step is worth 30 CU
Cumulative
30
Programme total
90
30 CU of 90 CU completed by the end of this step (33%)
Step 3, choose a course
Pick a course to see the detail
Credits, assessment, prerequisites and what each course unlocks later.
The twelve Forward capabilities
What this programme develops in you
Every Forward degree develops the same twelve cross-cutting capabilities. The profile below is the level this programme is designed to develop and verify by graduation, on the four-level scale in Part B2 of the Curriculum Compendium. Every claim above Level 2 is verified by someone other than the teaching lecturer.
- C1Intelligent Systems Fluency CriticalLevel 4 of 4, Leading
- C2Quantitative and Evidential Reasoning CriticalLevel 4 of 4, Leading
- C3Ethical and Contextual JudgementLevel 3 of 4, Independent
- C4Communication and PersuasionLevel 4 of 4, Leading
- C5Disciplinary Mastery CriticalLevel 4 of 4, Leading
- C6Technical Production and CraftLevel 4 of 4, Leading
- C7Problem Framing and Systems ThinkingLevel 4 of 4, Leading
- C8Collaboration and Multidisciplinary TeamingLevel 3 of 4, Independent
- C9Enterprise, Value and Commercial LiteracyLevel 3 of 4, Independent
- C10Professional Conduct and Workplace PerformanceLevel 4 of 4, Leading
- C11Learning to Learn and Adaptive Capacity CriticalLevel 4 of 4, Leading
- C12Stewardship and Public ContributionLevel 3 of 4, Independent
Critical capabilities for this programme: Intelligent Systems Fluency · Quantitative and Evidential Reasoning · Disciplinary Mastery · Learning to Learn and Adaptive Capacity. A graduate cannot pass out of this programme below the stated level on any capability marked critical.
Programme Learning Outcomes
What you will be able to do
Each outcome is assessed, and each is tagged with the Forward capabilities it is verified against.
- PLO 1Read, reproduce and critique current research literature in machine learning.
- PLO 2Train, adapt and align models at meaningful scale under real compute and data constraints.
- PLO 3Design valid evaluation regimes and identify the failure modes of published benchmarks.
- PLO 4Conduct and report original research to publication standard.
- PLO 5Build model capability for African languages, data and deployment conditions.
- PLO 6Assess and mitigate the safety, alignment and societal risks of the systems built.
- PLO 7Translate a research result into a deployable artefact with a costed path to use.
Why this programme exists
A research intensive degree for the small number of people who will build, adapt and evaluate frontier-class models on and for the continent. It exists because Africa is acquiring accelerated computing faster than it is acquiring the people qualified to use it well, and because the models the continent most needs — multilingual, low resource, efficient, domain specific — are not on anyone else's roadmap. The programme is deliberately small, deeply resourced and highly selective.
Admission requirements
A Bachelor's degree of at least Second Class Upper Division in computer science, artificial intelligence, mathematics, statistics, physics or a cognate quantitative discipline, with demonstrated programming ability; or a Second Class Lower with a substantial portfolio of built systems. All applicants complete a technical assessment and interview. Alternative pathways: Foundation Year, diploma entry with advanced standing, mature age entry and recognition of prior learning, all governed by Part A5 and Part G. Online applicants additionally complete the compulsory Digital Readiness orientation under Part D6.