Resilience under complexity
Research, test, fail, fix, document, try again. The reflexes that separate working engineers from people who took a course.
European Bachelor
Machine learning Engineering
- FEDE-Accredited
- 26 months
- 180 ECTS
- +5,000 coding hours
The credential
A real degree. Real engineering. Built for what's next.
Most tech training ends with a certificate. Holberton takes you further: a FEDE European Bachelor at EQF Level 6, designed to give your skills academic weight and international readability.
You graduate with more than proof of attendance. You leave with a recognised qualification, real engineering experience, and the practical skills employers are actively looking for.
Curriculum
Three periods. One bachelor.
Three intensive periods. Each builds on the last.
Systems & Tooling
Period 1 · Engineering foundations
From your first line of C to your first deployed pipeline.
Master the foundations every working engineer is expected to know — the command line, version control, low-level programming, and the mental model of how a computer actually runs your code.
Learning Objectives
- Navigate Linux and the shell with confidence — files, permissions, redirection, scripting.
- Write programs in C, manage memory by hand, and debug at the pointer level.
- Collaborate through Git and GitHub — branches, merges, conflicts, pull requests.
- Build modular C programs that compile cleanly and pass strict style checks.
- Manipulate core data structures: linked lists, stacks, queues, trees.
Technologies
- C
- GCC
- Linux
- Bash
- Git
- GitHub
- Make
- Valgrind
Key Projects
-
_printf
Reimplement C's printf from scratch, in pairs.
-
Shell
Build your own Unix shell that executes commands and pipes.
-
AirBnB clone (console)
Command-line storage engine and object model.
Full-Stack Web Architecture
Period 1 · Engineering foundations
Design and build secure, end-to-end web applications.
Architect software the way a real engineering team would. Model the system, build the backend, secure the data, and serve it through a working web client.
Learning Objectives
- Model software systems with UML — package, class, and sequence diagrams.
- Build secure REST APIs in Python and Flask with JWT authentication and role-based access.
- Design relational schemas and persist data through SQLAlchemy.
- Develop standards-compliant web clients that consume APIs and manage sessions.
- Diagnose multi-language defects, write tests, and document code professionally.
Technologies
- Python
- Flask
- SQLAlchemy
- MySQL
- JWT
- HTML5
- CSS3
- JavaScript
- UML
Key Projects
-
AirBnB clone (web)
Full-stack rebuild: API, database, authentication, frontend.
-
HBnB Evolution
Architect and document a layered system with UML before writing code.
Resilient Delivery & Infrastructure
Period 1 · Engineering foundations
Ship software that survives contact with production.
Take the leap from 'it works on my machine' to 'it runs reliably for thousands of users.' Containers, networking, deployment, and the operational thinking that makes software resilient.
Learning Objectives
- Type complex backends with TypeScript and build maintainable Node.js services.
- Coordinate asynchronous workflows with Promises and concurrency patterns.
- Containerize applications with Docker and orchestrate them with Compose.
- Reason about web infrastructure — DNS, load balancers, redundancy, monitoring.
- Present and defend a complete software delivery as a portfolio-grade project.
Technologies
- TypeScript
- Node.js
- Express
- Docker
- Docker Compose
- Nginx
- Linux Networking
Key Projects
-
Pagination engine
Deletion-resilient pagination for production-scale datasets.
-
Containerized multi-tier app
Reverse proxy, load-balanced services, full Compose stack.
-
End-of-period portfolio
Documented, deployed, and presented to a jury.
AI-Augmented Engineering & ML Foundations
Period 2 · Machine learning and AI
The math, the tools, and the AI workflow that underpin every model.
Build the engineering muscle that machine learning sits on top of. Numerical computation, statistics, and structured AI-assisted workflows that compound your productivity.
Learning Objectives
- Engineer numerical pipelines with NumPy — broadcasting, vectorization, matrix operations.
- Apply probability and statistics to quantify uncertainty in models.
- Architect monolith vs microservices APIs and document them with OpenAPI specifications.
- Use AI assistants strategically — prompt libraries, IDE automation, productivity baselines.
- Train neural networks responsibly with regularization, dropout, and early stopping.
Technologies
- Python
- NumPy
- pandas
- Matplotlib
- OpenAPI
- AI IDE Tooling
Key Projects
-
Vector & matrix engine
Reimplement core linear algebra operations from scratch.
-
Probability distributions library
Build PDFs, CDFs, and samplers without scipy.
-
Architecture blueprint
Compare monolith and microservices designs for the same product.
Deep Learning & Computer Vision
Period 2 · Machine learning and AI
Design, train, and deploy modern neural networks.
Move from foundational ML into the deep learning systems that power production AI. CNNs, transfer learning, generative models, and the optimization techniques that make them work.
Learning Objectives
- Architect deep convolutional networks in TensorFlow and Keras.
- Apply transfer learning and fine-tuning to hit performance targets faster.
- Build object detection inference pipelines: decode, NMS, mAP-ready predictions.
- Train and stabilize generative adversarial networks (GANs and WGAN-GP).
- Optimize models with Bayesian search and dimensionality reduction.
Technologies
- TensorFlow
- Keras
- scikit-learn
- OpenCV
- PCA
- Bayesian Optimization
Key Projects
-
Image classifier
CNN trained end-to-end on a real-world dataset.
-
Object detector
Decode predictions, apply non-max suppression, evaluate with mAP.
-
GAN training pipeline
Stabilize adversarial training and generate synthetic samples.
NLP, Reinforcement Learning & European Modules
Period 2 · Machine learning and AI
Language, decisions, and the European credential.
Bring AI into language and behavior. Build NLP systems and reinforcement learning agents — and earn the European credentials that make this a recognized bachelor.
Learning Objectives
- Engineer text representations: TF-IDF, word embeddings, transformer tokenization.
- Train and deploy transformer-based seq2seq and question-answering systems.
- Implement reinforcement learning agents: Q-learning, DQN, REINFORCE.
- Build reliable data pipelines: API ingestion, validation, storage, ML-ready datasets.
- Achieve A2 proficiency in a second European language and complete the European citizenship module.
Technologies
- Hugging Face
- Transformers
- Gym
- keras-rl2
- NLTK
- SQL
- NoSQL
Key Projects
-
QA assistant
Console chatbot powered by a pretrained transformer.
-
DQN agent
Train a deep reinforcement learning agent to solve a control task.
-
Data pipeline
End-to-end ingestion from API to ML-ready dataset.
Algorithmic Mastery
Period 3 · Mastery and portfolio
The algorithms and data structures that come up in every senior interview.
Sharpen the algorithmic thinking that separates a junior who can ship from a junior who gets hired. Big-O reasoning, advanced data structures, and the discipline to deliver under constraints.
Learning Objectives
- Select and justify algorithmic paradigms: greedy, divide-and-conquer, backtracking.
- Implement and manipulate advanced data structures: AVL trees, heaps, skip lists.
- Analyze time and space complexity with Big-O notation and refactor for performance.
- Deliver production-grade code under explicit constraints (restricted calls, memory bounds).
- Defend algorithmic decisions in technical interview-style oral exams.
Technologies
- C
- Python
- Node.js
- Big-O Analysis
- Algorithm Design
Key Projects
-
Advanced data structures
Build AVL trees, heaps, and skip lists in C.
-
Algorithmic challenge set
Solve interview-grade problems under timed constraints.
-
Mock technical interview
Defend two random projects orally to a panel.
Applied AI Project & Portfolio Defence
Period 3 · Mastery and portfolio
Conceive, build, defend — your end-to-end AI MVP.
The capstone. Turn a loose idea into a working AI product, deployed and documented. Defend it to a jury. This is the project that goes on your CV and into every job interview after graduation.
Learning Objectives
- Scope an AI MVP from a loose theme — users, success criteria, delivery plan.
- Design a compliant data and privacy approach with documented controls.
- Evaluate ethical risks: bias, misuse, safety failure modes, and concrete mitigations.
- Implement an end-to-end pipeline: data ingestion, training, inference, deployment.
- Communicate results — GitHub README, blog post, slide deck, live demo.
Technologies
- Python
- TensorFlow
- Docker
- Git
- Project Management Tools
- Cloud Deployment
Key Projects
-
AI MVP capstone
A real, deployed AI product of your choosing, defended to a jury.
Professional Practice & European Modules
Period 3 · Mastery and portfolio
From graduate to hire.
The bridge between the bachelor and your first role. Workplace practice, intercultural management, advanced language proficiency — and the professional habits employers screen for.
Learning Objectives
- Reach B1 proficiency in a second European language, in writing and speaking.
- Apply intercultural management principles to real workplace scenarios.
- Maintain professional conduct, attendance, and digital security across the program.
- Deliver original work and meet program milestones reliably.
- Communicate technical decisions to non-technical stakeholders.
Technologies
- English (Required)
- Second European Language
- Professional Workflows
Key Projects
-
Language proficiency exam (B1)
Written and oral, in your chosen European language.
-
Intercultural management assessment
Online MCQ on the FEDE reference framework.
-
Continuous professional review
80%+ professional score across the period.
How you learn
Train in the conditions of the job. Not in theory.
Software engineering is fast-paced, demanding, and constantly evolving. We train you for that reality from day one — not after.
Real autonomy
You learn to mobilize resources, structure your approach, ask for help effectively, and progress without waiting for a lecture.
Professional reflexes
Working with constraints. Delivering outcomes. Maintaining quality. Communicating technical choices. The job, not the simulation.
Adapting to change
Exposure to varied problems and contexts. You learn to evolve with tools and methods, not depend on a single technology.
Outcomes
Built for the job. Ready for the role.
Roles our graduates apply for:
- Machine Learning Engineer
- AI Engineer
- MLOps Engineer
- AI/ML Software Engineer
- Computer Vision Engineer
- NLP Engineer
- Data Scientist
How to join
Two paths in. One bachelor out.
We accept students through two routes. Both lead to the same European Bachelor.
Choose the one that fits your background.
You're ready if…
- You're 16+
- You can study full-time in English
- You're ready to commit two intensive years.
- You're more curious than credentialed.
- You want a degree and the skills to use it.
Path 1
Standard admissionApplicant with HSC, Baccalauréat, or equivalent minimum
- Apply online
- Motivation interview
- Welcome pack & enrollment
- Day one: you build
Path 2
Entry testApplicant without HSC or Baccalauréat
- Apply online
- Entry test (logic, problem-solving, basic programming)
- Motivation interview
- Welcome pack & enrollment
- Day one: you build
Start your application.
Tell us a bit about you. We respond within 48 hours.
Thanks — we'll be in touch.
We typically respond within 48 hours. Or call us directly: +230 5255 0390.
Questions
Things parents and applicants ask.
Is the Bachelor taught in English?
Yes. All instruction, projects, and materials are in English. We require a B2 minimum English level (C1 is comfortable) — this prepares you for the international tech industry, where English is the working language.
Is the European Bachelor recognized in Mauritius?
Yes. The degree is awarded by FEDE (Fédération pour l'Éducation en Europe), positioned at EQF Level 6 and internationally recognized. We can provide formal documentation on request.
How does this compare to a Mauritian university degree?
Different model. Local universities focus on lectures, exams, and theory across three or four years. Holberton focuses on projects, peer collaboration, and shipping real software across two intensive years. The credential is European, the degree is at the same EQF level (Level 6), and the practical experience is incomparable.
What can my child do after the Bachelor?
Two paths: enter the workforce as a junior engineer (most graduates worldwide), or continue to a Master's program in Europe or Mauritius using the 180 ECTS earned.
What does a typical week look like?
Five days a week on campus, 8:30 to 15:00 — project work, peer collaboration, and code review. After campus, students continue building independently. The program is full-time and intensive.
What's the campus like?
Our campus is at 17 Lislet Geoffroy Street in Port Louis. Modern facilities built for project-based learning. We welcome on-site visits — book one through our contact form.
What if my child doesn't have a tech background?
Most Holberton students don't, when they start. The Bachelor is built for committed beginners as much as for experienced builders. What matters is curiosity and the willingness to build.
What does the application timeline look like?
Apply any time. We respond within 48 hours, then schedule the motivation interview (and entry test, if applicable) within 1–2 weeks. From application to enrollment is typically 3–4 weeks. Cohort start dates are published on the application form.