Delft University of Technology
Top university
Just Landed
New Today
PhD in Decentralized and Trustworthy AI Pipelines (EU project WALTZ) Delft University of Technology in Netherlands
Degree Level
PhD
Field of study
Computer Science
Funding
PhD employment for 4 years in principle, split into an initial 1.5-year contract and a subsequent 2.5-year contract if progress is satisfactory. Salary ranges from €3204 to €4051 gross per month based on full-time employment (38 hours), plus 8% holiday allowance and an 8.3% end-of-year bonus. TU Delft also offers a customizable compensation package, health insurance discounts, and a monthly work costs contribution.
Deadline
Sep 24, 2026
Country
Netherlands
University
Delft University of Technology

How do I apply for this?
Sign in for free to reveal details, requirements, and source links.
Apply for this position
Keywords
Suggested positions
About this position
PhD position at Delft University of Technology (TU Delft) in the Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS), within the Distributed Systems group of the Department of Software Technology.
The project is part of the EU Horizon Europe Innovation Action WALTZ (Workflow-Driven AI-Enabled Lawful and Trusted Data Spaces for Public Authorities). The research focuses on decentralized AI pipelines, trustworthy AI, distributed systems, federated learning, privacy-preserving machine learning, and LLM-based systems. The goal is to build training and inference pipelines that can operate across administrative boundaries without a central data aggregator, while remaining robust to Byzantine participants, synthetic-data poisoning, privacy leakage, and unreliable model outputs.
Possible research directions include combining real and synthetic data under controlled mixing strategies, studying robustness against adversarial behavior, quantifying membership inference and reconstruction risks, evaluating differential privacy trade-offs, and designing quorum or semantic-agreement mechanisms for distributed LLM inference. The evaluation side includes benchmarks, cross-validation schemes, reporting protocols, and open reproducible software for accuracy, robustness, bias, generalisation, and privacy leakage.
You will be supervised by Dr. Jérémie Decouchant. The role includes collaboration with partners across Europe and real-world public authorities, publication at top venues, and a modest teaching/supervision load of up to 15%.
Funding and conditions: 4-year PhD employment in principle, with salary from €3204 to €4051 gross per month, plus holiday allowance and end-of-year bonus. TU Delft also mentions health insurance discounts, a monthly work costs contribution, and relocation support to the Netherlands.
Eligibility highlights: MSc in Computer Science, Data Science, Artificial Intelligence, Electrical Engineering, or a closely related field; strong Python and PyTorch skills; experience with Linux-based GPU/HPC clusters; and strong interest in trustworthy AI, privacy, robustness, and reproducibility. Excellent English is required.
Application deadline: 24 Sep 2026. Apply online and upload a cover letter, CV, transcripts, thesis if applicable, and proof of English proficiency if applicable.
Funding details
PhD employment for 4 years in principle, split into an initial 1.5-year contract and a subsequent 2.5-year contract if progress is satisfactory. Salary ranges from €3204 to €4051 gross per month based on full-time employment (38 hours), plus 8% holiday allowance and an 8.3% end-of-year bonus. TU Delft also offers a customizable compensation package, health insurance discounts, and a monthly work costs contribution.
What's required
An MSc completed or near completion in Computer Science, Data Science, Artificial Intelligence, Electrical Engineering, or a closely related field. Applicants should have a solid foundation in machine learning and/or distributed systems, strong Python programming skills, experience with a modern deep learning framework such as PyTorch, and hands-on work on Linux-based GPU/HPC clusters. Familiarity with federated or decentralized learning, fault tolerance, generative models, privacy-preserving ML, or LLM-based systems is a strong asset. Candidates should show interest in trustworthy AI, privacy, robustness, adversarial behaviour, evaluation methodology, reproducibility, strong analytical skills, independence to carry a four-year research agenda, excellent English, collaborative ability, and commitment to open science.
How to apply
Apply online via the application button by 24 Sep 2026. Upload a cover letter, CV, BSc and MSc transcripts, MSc thesis if applicable, proof of English proficiency if applicable, and optionally code repository links. Do not apply by email or post.
More information can be found here
Ask ApplyKite AI

How do I apply for this?
Sign in for free to reveal details, requirements, and source links.