Join the lab

I am hiring a postdoctoral researcher at TU Delft on AI for tabular and temporal data, with a particular interest in foundation models and generative AI for quantum technologies.

The broader vision is to develop data-centric AI methods for complex scientific and engineering problems, including applications in quantum hardware and quantum data management — making AI a first-class tool across the quantum stack, while bringing the rigour of data systems to scientific AI.

Who I’m looking for

I would be very happy to hear from:

• candidates with a strong AI background — especially in foundation models, generative AI, or representation learning over tabular and temporal data; and
• candidates from quantum hardware, quantum error correction, or quantum calibration / control who are excited to work at the intersection of AI and quantum computing.

You don’t need to tick both boxes — curiosity about the other side is what matters most.

How to apply

If you are interested, please send me an email. There is no need for an exhaustive CV — a brief introduction and 1–3 selected papers are enough, ideally from top AI venues such as NeurIPS, ICML, ICLR, or strong Q1 journals.

Quantum error correction (QEC) is fast becoming a data problem: training AI decoders needs large, well-organised datasets of syndromes, measurements, and metadata. In this project you will build the infrastructure to collect, organise, and export QEC datasets for machine-learning models — extending an existing prototype with data ingestion, metadata management, and dataset discovery.

It is a hands-on mix of system building, data management, machine learning, and experimental evaluation, and a chance to help shape the data layer of a brand-new scientific AI field.

Good fit if you have a strong interest in data systems, solid Python, and some scientific-computing experience. Quantum expertise is not required.

AI-enhanced QEC decoders are appearing fast — but they are hard to compare. This project builds a benchmark repository and model framework for quantum error correction decoders, with the infrastructure to make results reproducible and comparable across the community.

You will design the benchmarking layer for a new scientific AI field that combines machine learning with quantum computing, putting real software-engineering care into reproducibility and fair comparison.

Good fit if you have strong Python and machine-learning skills. Quantum knowledge is helpful but not required.

We simulate quantum circuits inside a relational database, where tensor contractions become SQL joins and aggregations (see our Qymera work). This project uses AI and data-centric techniques to generate diverse training circuits and SQL workloads, then learns to simulate and compile circuits more efficiently.

You will work across databases, machine learning, and quantum computing — generating data, training models, and optimising query plans for circuit simulation.

Good fit if you are interested in databases, machine learning, or quantum computing, with strong Python; SQL knowledge is very helpful.

Can you retrieve a record from a database without revealing which record you asked for? This project develops software that connects classical databases to quantum-network private-query protocols, enabling secure record retrieval that hides the query target.

Co-supervised with Wolfgang Löffler (Leiden University), the work bridges database systems and quantum information — from the SQL side down to the quantum network.

Good fit if you have some background in quantum information (useful, not essential) and experience with Python, SQL, or systems programming.

A more protocol-focused companion to the Private Quantum Database project: you will study quantum protocols for private database queries over near-term quantum networks, analysing protocol design, quantum encodings, and hardware constraints.

Expect protocol comparison, numerical simulations, privacy analysis, and practical design rules for photonic-hardware implementation. Co-supervised with Wolfgang Löffler (Leiden University).

Good fit if you can program in Python or Julia. A background in quantum communication or cryptography helps but is optional.

An internship with Philips on agentic AI for service maintenance. You will build AI agents that autonomously search and reason over maintenance records and device logs to accelerate troubleshooting.

You will research tool-use and multi-step reasoning approaches, implement unified retrieval across heterogeneous data, and prototype and test with real maintenance data.

Good fit if you are pursuing a BSc, MSc, or PhD in CS, AI, or Data Science, with solid ML and NLP and strong Python; familiarity with PyTorch / TensorFlow and vector databases is a plus.

An internship with Philips applying generative AI to unstructured maintenance text — extracting structured information, classifying documents, and improving information retrieval.

You will evaluate state-of-the-art NLP methods, prototype with real service reports and technician notes, collaborate with domain experts, and present your findings to stakeholders.

Good fit if you are pursuing a BSc, MSc, or PhD in CS, AI, or Data Science, with a good grasp of ML and generative AI and proficiency in Python; familiarity with PyTorch, TensorFlow, or Hugging Face Transformers is a plus.

Don’t see a topic that fits? Propose your own. We especially welcome MSc students from DSAIT, Computer Science (CS), and QIST who want to shape a thesis at the intersection of data systems, AI, and quantum computing.

Bring an idea — or just your interests — and we will work with you to turn it into a concrete project across our two research directions: data systems for AI, and data systems for quantum computing.

Good fit if you are a TU Delft MSc student with curiosity and initiative. Send a short email about what excites you, and we’ll take it from there.