All internship topics (first and second year) must be validated by the directors. To do so, send an email to dir.msct.trai@polytechnique.fr mentioning the internship topic (including a half-page summary of the work to be done), duration, supervisor, and location. Once validated, please indicate Sonia Vanier (for M1 students) or Luiz Chamon (for M2 students) as the “Enseignant référent” (reference) at École polytechnique on the official “convention de stage” (internship contract). Note that all four directors of TRAI will be part of the jury during you internship defense.

Period: from the beginning of April to the end of August (minimum: 4 months)
Defense: First half of September

The internship can be carried out either at a private company or at an academic lab. The research component of the internship is not mandatory, but strongly encouraged. Students are requested to choose a topic related to the curriculum of their graduate degree program, i.e., involving ML/AI or trustworthiness and reliability aspects, ideally both. The topic can be theoretical (understanding of a theoretical result and extension to a new setting), applied (development and implementation of a solution), or a mixture of both.

Period: from the beginning of April to the end of September (minimum: 5 months)
Defense: First half of September (the internship can continue after the defense)

A research internship carried out either in the R&D department of a company or in a public research lab. The student must produce original work related to ML/AI and their trustworthiness aspects, ideally involving different topics covered during the program. This work can be theoretical or applied so long as it contains novel ideas developed and validated by the student.

After the internship, the student must hand-in a report ranging from 15 to 20 pages, figures and references included. It must containa general presentation of the topic and a description of the state-of-the-art in the field (theory and/or methods). The contributions of the student must be clearly identified and explained in details. There is no need for an exhaustive description of all codes produced during the internship, though algorithms highlighting the challenging tasks solved by the student should be presented and explained in the report (if need be, relevant parts of the code can be included as an appendix).

Below we provide more detailed, section-by-section guidelines:

• Introduction: motivation (why is your problem important? why is it challenging?) and a clear description of your goals;
• Related work: description of relevant existing theory and methods used to solve the same or related problems as well as methods used to inspire your solution;
• Proposed method (one or several sections): technical details about your scientific contributions and the solution you developed;
• Experimental results: test and validation of the proposed method, including a discussion of pros and cons of your solution (this section may not be relevant for theory-heavy internships);
• Conclusions: what did you learn, ideas about future work…
• References: relevant bibliography (include the final version of the papers you used, i.e., always cite the journal/conference versions rather than the arXiv, if available.

Oral examinations will consist of a 15 minute (sharp!) talk on the content of the report. It is followed by 10 minutes of questions from the jury.

Kindly provided by AI-ViC students (2024-2025)

• Start very very early, beginning of October preferably
• I handed out my CV at the fair and spoke for some time to the recruiter, and they contacted me later. Tips: Get as many details as you can about the company, projects and interview process at the fair, take notes, prepare accordingly and bring up the details you found interesting during the interviews.
• Talk to as many people as you can (reduce uncertainty and put yourself on the map)
• To strengthen internship pipelines, we should tap into our alumni network and boost visibility between former interns and incoming students. For instance, after my contribution at Amazon, my manager is eager to hire SDE interns straight from our university’s next cohort. By organizing regular meet-and-greet sessions or panel discussions where recent graduates share their experiences with current students, we create direct referral channels and give hiring managers tangible reasons to scout talent from our program.
• Use a centralised tracker such as https://github.com/henriChevreux/FR_jobspy.git
• If you want to go to big tech, apply on the first day the opening opens. Contact people from the company before applying. Check that the company / lab is a fit for you — not only that you are a fit for them.
• Work on a niche project and reach out to people who would care about that niche.
• Start as soon as possible.
• Don't be afraid to reach out to people directly.
• LinkedIn Outreach and Networking events
• X-forum is a good chance – If you try more, you have more – Networking is all you need.

For M1 students: Access internship proposals from the Polytechnique engineering student page.

If you would like to see more internship proposals preselected for their decent research components, please follow this procedure:

• Go to https://stages.dix.polytechnique.fr/home
• Register as student, with your official email address, and select the program “CSC_52994 - 202X”
• Damien Rohmer will validate your registration
• Once validated, they can connect with:
  • login: your email address
  • password: 3AstudentDIX

• CNIL: Auditing compliance of AI Systems
  • Description: CNIL is the French Data Protection Authority and will have a central role in the application of the EU AI Act that is progressively entering into force. In the AI department, you will support the effort to put in place CNIL's technical strategy for auditing AI systems. This is a unique opportunity to participate in the very beginning of AI regulation.
  • Keywords: AI Audit, AI Act, AI Regulation, AI Ethics
  • Theory/applied: In-between research and operational missions
  • Contact: Nicolas Berkouk nberkouk@cnil.fr