Master’s in Privacy, Federated and Graph Engineering
Online master's programme of 1,900 hours in privacy engineering, federated learning and graphs. 12 modules covering privacy in data systems, anonymisation, differential privacy, cryptography for learning, federated learning, graph theory and databases, graph machine learning, privacy attacks and auditing, synthetic data and regulation, with an assigned technical tutor.
Take the controls.
Run a real engineering trade-off from this field. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Assess re-identification risks in datasets.
Apply anonymisation and differential privacy techniques.
Use cryptographic techniques for secure learning.
Train models through federated learning.
Develop machine learning models on graphs.
Audit privacy leakage in trained models.
Integrate GDPR requirements into data and AI projects.
Who it's for
- Privacy engineers building privacy-enhancing technologies into data pipelines.
- Federated learning engineers training models across organisations.
- Graph data scientists working with networks and knowledge graphs.
- Technical data protection officers who need hands-on command of PETs.
- Trustworthy AI researchers focusing on leakage, auditing and unlearning.
- Data sharing architects designing data spaces and clean rooms.
- Engineering or science graduates aiming to enter the healthcare sector.
What you'll work with
- Regulation (EU) 2016/679 (GDPR)
- ISO/IEC 27701 (privacy information management)
- ISO/IEC 29100 (privacy framework)
- ISO/IEC 20889 (privacy-enhancing data de-identification techniques)
- ISO/IEC 27559 (privacy-enhancing data de-identification framework)
- Regulation (EU) 2022/868 (Data Governance Act)
- Regulation (EU) 2025/327 (European Health Data Space)
- Regulation (EU) 2024/1689 (AI Act)
- Flower
- OpenDP
- ARX
- PyTorch Geometric
- Neo4j
12 modules · 1900 h
1.1Concepts of privacy and identifiability1.2Personal, pseudonymised and anonymous data1.3Re-identification and linkage attacks1.4Privacy by design principles1.5Privacy framework according to ISO/IEC 291001.6Privacy risks in machine learning1.7Graph data and its sensitivity1.8Privacy-enhancing technologies1.9Map of techniques by use case1.10Case study: risks of a customer dataset
2.1The k-anonymity model2.2l-diversity and t-closeness2.3Generalisation and suppression2.4Record microaggregation2.5Anonymisation of time series and trajectories2.6Anonymisation of social graphs2.7Measuring utility after anonymisation2.8Criteria of European authorities2.9Anonymisation tools2.10Lab: k-anonymity with ARX
3.1Formal definition and privacy budget3.2Laplace and Gaussian mechanisms3.3Composition and accounting3.4Local differential privacy3.5Private queries and statistics3.6DP-SGD for neural networks3.7Differential privacy on graphs3.8Choosing epsilon in practice3.9Real-world differential privacy deployments3.10Practical exercise: private statistics with OpenDP
4.1Partially and fully homomorphic encryption4.2Secure multi-party computation4.3Secret sharing4.4Trusted execution environments4.5Zero-knowledge proofs4.6Private set intersection4.7Computational cost of each technique4.8Available cryptographic libraries4.9Combining techniques4.10Lab: encrypted inference with homomorphic encryption
5.1Federated client-server architecture5.2The FedAvg averaging algorithm5.3Horizontal and vertical federated learning5.4Non-IID data and heterogeneity5.5Communication efficiency5.6Secure aggregation protocols5.7Federated learning with differential privacy5.8Poisoning attacks and defences5.9Flower and NVIDIA FLARE frameworks5.10Workshop: federated training with Flower
6.1Graph theory and representations6.2Centrality measures6.3Community detection6.4Paths and connectivity6.5Directed, weighted and heterogeneous graphs6.6Dynamic and temporal graphs6.7Graph databases6.8Cypher and GQL languages6.9Knowledge graphs6.10Practical exercise: network analysis with NetworkX and Neo4j
7.1Node embeddings with node2vec7.2Graph convolutional networks7.3GraphSAGE and sampling7.4Graph attention networks7.5Link prediction7.6Graph classification7.7Typed heterogeneous graphs7.8Scaling to large graphs7.9PyTorch Geometric and DGL7.10Lab: node classification with GNNs
8.1Inference attacks on GNNs8.2Membership and link inference8.3GNNs with differential privacy8.4Federated graph learning8.5Subgraphs distributed across organisations8.6Federated knowledge graphs8.7Unlearning in graph models8.8Leakage assessment8.9Cases in finance and healthcare8.10Practical exercise: federated GNN on subgraphs
9.1Membership inference9.2Model inversion9.3Training data extraction9.4Leakage in language models9.5Empirical auditing of differential privacy9.6Machine unlearning techniques9.7Memorisation in models9.8Auditing tools9.9Risk report9.10Workshop: membership inference attack and mitigation
10.1Synthetic data generation10.2Synthetic data with differential privacy10.3Evaluating fidelity and privacy10.4Synthetic graph generation10.5European data spaces10.6Data clean rooms10.7Data sharing governance10.8Data contracts and agreements10.9Cases in healthcare and banking10.10Lab: private synthetic data with SDV
11.1GDPR and privacy-enhancing technologies11.2Data protection impact assessments11.3International data transfers11.4Data Governance Act11.5European Health Data Space11.6The AI Act and data governance11.7ISO/IEC 27701 and 27559 standards11.8Auditing and certification11.9Accountability of the controller11.10Case study: DPIA for a federated learning project
12.1Definition of the sensitive use case12.2Privacy risk analysis12.3Selection of protection techniques12.4Architecture design12.5Prototype implementation12.6Utility and privacy evaluation12.7Regulatory compliance analysis12.8Conclusions on utility and privacy12.9Master's Thesis report12.10Master's Thesis: defence before a panel
How you'll learn
100% online training on the SEIUM campus, with downloadable material, recorded classes and an assigned technical tutor. There are no fixed timetables: you progress at your own pace within the intake, so you can combine it with your job.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is Flower, OpenDP, ARX, PyTorch Geometric and Neo4j, in free, academic or evaluation versions. Practical work is based on real procedures: anonymisation of personal datasets, federated training across organisations and leakage auditing of graph neural networks. Assessment is continuous by module, plus the final Master's Thesis.
Standards covered: Regulation (EU) 2016/679, ISO/IEC 27701, ISO/IEC 29100, ISO/IEC 20889, ISO/IEC 27559, Regulation (EU) 2022/868, Regulation (EU) 2025/327, Regulation (EU) 2024/1689.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Where this takes you
Requirements
- A university degree in engineering, science or architecture, or equivalent demonstrable professional experience.
- An internet connection and a computer capable of running the programme's software.
- English sufficient to read technical standards.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 1,900 hours over 12–14 months. You set your own pace within each intake, and your tutor helps you plan the workload around your job.
You work with Flower, OpenDP, ARX, PyTorch Geometric and Neo4j, using free, academic or evaluation versions. Your tutor indicates the free alternative wherever one exists.
A SEIUM Institute diploma, awarded on passing the continuous assessment and the Master's Thesis. It is SEIUM's own qualification, not an official state degree; its recognition depends on the programme completed and the regulations of each country.
Yes. Modules passed in a course or postgraduate diploma in the same area are recognised in the higher programme. Ask about the recognition policy when you enrol.
A university degree in engineering, science or architecture (or equivalent professional experience), a computer able to run the programme's software and English sufficient to read technical standards.
Start dates are set by SEIUM's faculty for each intake. Our admissions team will confirm the next available date when you apply.
Related programmes
Master’s in Real-Time Analytics and Stream Processing Engineering
Online master's programme of 1,900 hours in real-time analytics and stream processing: 12 modules covering event stream processing, distributed messaging, Apache Flink, streaming engines, real-time analytics,…
Master’s in Reinforcement Learning, Control and Safe AI Engineering
Online master's programme of 1,900 hours in reinforcement learning, control and safe AI engineering. 12 modules covering sequential decision-making, deep RL, model-based learning, classical and modern…
Master’s in Optimisation under Uncertainty Engineering
Online master's programme of 1,900 hours in optimisation under uncertainty: 12 modules covering uncertainty in decision-making, scenario models, two-stage and multi-stage stochastic programming, chance constraints, robust…
More in Data, AI, Simulation & Analytics
All Data, AI, Simulation & Analytics programmes → · How to apply · Fees & financing · Scholarships