Prospective students and postdocs
I welcome applications from outstanding prospective PhD students and postdoctoral researchers interested in building reliable learning systems that couple expressive models with constraints, dynamics, and mechanistic structures.
Why join this group?
This is an opportunity to join a growing research program at a stage where new researchers can genuinely help shape its scientific direction. We develop machine-learning methods that make better use of structure—such as constraints, dynamics and simulations, with the aim of building methods that are mathematically principled, computationally effective, and useful in realistic settings.
PhD students and postdocs are encouraged to develop substantial intellectual ownership of their work: identifying important questions, developing the technical ideas and experiments needed to address them, and positioning the resulting work within the broader machine-learning community.
The group is small enough for close scientific interaction and visible individual contributions, while being embedded in the broader research environment of Inria, Université Grenoble Alpes, and THOTH. This provides access to expertise and collaborations across machine learning, optimization, and scientific applications, while leaving room to develop an independent research identity.
Research directions and candidate fit
Current research directions include:
- learning with implicit constraints, optimality conditions, and dynamical systems;
- learning from simulations and synthetic data;
- representation and self-supervised learning.
This is a particularly good fit for candidates who enjoy moving between mathematical questions and careful empirical work. Strong preparation in machine learning, probability, optimization, statistics, or a neighbouring quantitative field, together with hands-on experience with modern deep-learning methods and tools, is essential.
Mentoring and group culture
I aim to offer serious scientific engagement, candid and constructive feedback, increasing research independence, fair credit, and a professional environment where people can disagree thoughtfully. The group is ambitious about the quality of its research while respecting reasonable working boundaries.
The group guide gives an introduction to the research culture and links to the detailed advising and research-practice documents. It is intended to help you decide whether the group is a good mutual fit.
Current openings
All advertised PhD and postdoctoral positions are listed through the official Inria recruitment platform. To find positions connected to the group, filter the listings by “Thoth.” The official posting is the source of truth for eligibility, deadlines, and application materials.
If you are interested in an advertised position:
- Please apply directly through the corresponding Inria job posting.
- Please do not send me a separate email about an advertised application. To keep the process fair and manageable, applications are reviewed through the official system and shortlisted candidates will be contacted directly.
- In your application, make the research fit concrete: briefly explain which questions on this site interest you, alongside your relevant coursework, research, and technical experience.
- Before an interview, shortlisted candidates should read the group guide.
- Interviews assess technical foundations, research judgment, and mutual fit. Interviews also include a conversation about how we might work together.
Future opportunities
If no suitable position is currently listed and you would like to make a focused enquiry about a future opportunity, a concise message may include:
- a CV highlighting strong foundations in machine learning, probability, optimization, or statistics, together with hands-on experience with modern deep-learning methods and tools;
- a short explanation of one or two specific research questions or papers on this site that connect to your background, and the direction you would like to explore;
- a link or attachment to a relevant thesis, research paper, or substantial technical project, with a brief explanation of your own contribution;
- for prospective PhD students, academic transcripts; for postdoctoral candidates, your most relevant research contributions and evidence of independence.
I receive many enquiries and cannot guarantee an individual reply to every one. A clear, specific fit is the best way to make an enquiry useful.