Internship at École polytechnique: from the company placement to research

Three internships, three different purposes

At École polytechnique, internships are compulsory and assessed. They can take place in a company, in academic research or in research and development, in associations or in the military, in France or abroad. Each ends with a detailed report and a defence before the School. According to the official description of the cycle, this time in the field makes up about a third of the course.

Yet you should avoid speaking of “an internship” in the singular. The second-year placement resembles neither in length nor in purpose the third-year research internship, which is itself distinct from the final-year project. Starting by identifying which one concerns you stops you looking in the wrong place, with the wrong question.

• Second year: a company internship, to observe an organisation.
• Third year: a research internship, to work on an open problem.
• End of the course: an engineer-level or early-career researcher project lasting four to six months.

The second-year company internship

This internship takes place in the summer, between mid-June and the end of August, with a minimum of 12 weeks (about three months). It has three aims: to put yourself in a job-hunting situation, to observe how an organisation and its day-to-day management work, and to carry out a piece of work on a specific operational assignment.

The range of host organisations is deliberately limited: a company, a start-up or a public administration with more than five employees. A laboratory, a research centre, an association or a hospital does not qualify for this internship, which may surprise a student drawn to research. The constraint is logical: this is the time to understand how an organisation works, not to do science.

A good second-year assignment is defined tightly enough to produce a deliverable: analysis of a process, building a dashboard, a market study, a prototype. A topic that is too vague (“discover the profession”) does not allow you to write a solid report.

The third-year research internship

Here the goal changes: to solve an open problem using the scientific method, to compare several approaches critically and to learn to manage the uncertainty that comes with research. The internship runs from the end of March to the end of August and lasts at least 16 weeks; the School mentions a range of four to six months.

Proposals must be approved by the School’s teaching and research departments, so you need to talk early with an academic adviser. The student does not simply choose a laboratory: they choose a topic, a supervisor and a team, and have the whole package approved.

What to check before committing

• Is the question really open, or already answered in the literature?
• How much time will the supervisor have available?
• What resources (computing, instruments, data) will be accessible from the first month?
• Is the laboratory used to hosting interns and letting them contribute to a publication or an internal report?

Internship forms and agreement: a multi-step process

The administrative side is more formal than you might imagine, and it follows precise calendars depending on the type of internship.

For the company internship

The student fills in an internship form on the dedicated tool, has it validated by email by the supervisor and by the organisation’s human resources department, then hands it to their head of teaching with their insurance certificates: civil liability and, outside the European Union, repatriation and health cover, or the European Health Insurance Card within the Union. The international relations and internships office then issues the agreement once the form is approved, and several signatories receive an electronic signature request.

For the research internship

Two stages: an internship project form, signed by the supervising academic and handed to the internships office by the second Friday of December at the latest; then the final internship form, signed by the academic responsible for the option, to be handed in before the second Friday of February.

Above all, remember that a late signature blocks the issuing of the agreement. Warn your host organisation about these deadlines from the very start of the discussion.

Where to find offers and how to apply

Offers are available on the School’s careers support platform and on the alumni community’s job board. Organisations can also submit a proposal, either as part of a partnership with the School or through the careers platform, which then validates publication.

Applying methodically means using a variety of channels:

• offers published on the School’s platforms;
• direct contact with researchers whose work interests you, after reading their recent publications;
• the alumni network, useful above all for company and start-up internships;
• fairs and meetings organised by the School with employers.

For a research internship, write a short message that names a paper from the laboratory, asks a precise question and states your availability. A generic email sent to twenty laboratories almost always goes unanswered.

Present an approach rather than a promised result

At interview, avoid announcing that you will “solve” a problem. Instead, describe what you have already practised (methods, tools, course projects), what you want to explore further and how you will organise your work: reading, first experiments, progress meetings.

Scientific work can end in limits or negative results, which remain useful if they are well documented. A supervisor appreciates a student who can write “this avenue does not work, and here is why”, with evidence to back it up.

Plan regular exchanges from the outset, for example a weekly meeting, and keep a laboratory notebook or logbook. This also makes the final report easier, since it must distinguish your own contributions from the collective work of the team or the company.

Illustrative example: choosing between two research offers

Imagine, with rounded figures, a third-year student who receives two replies for her research internship.

• Offer A: a public laboratory, supervised by a research director available two hours a week, an open topic, a 20-week framework.
• Offer B: the R&D department of an industrial group, supervised by a senior engineer, a more applied topic, an 18-week framework.

She compares the clarity of the question, the supervisor’s availability, access to data and the chance to present her results. She also consults her academic adviser before deciding, because the proposal has to be approved by the School. This example is only an illustration: real conditions depend on each organisation and on the option followed.

Common mistakes to avoid

• Mixing up the periods: looking for a laboratory for the second-year internship when that internship is aimed at an organisation with more than five employees.
• Waiting until February to start the research internship process, when the project form is due in December.
• Forgetting the insurance certificates, which are essential for the agreement to be issued, especially abroad.
• Accepting a topic without clear supervision, at the risk of spending weeks with no feedback.
• Promising results that no serious supervisor could guarantee.
• Neglecting language: for certain destinations, or for non-French-speaking students, proof of language level is required.

From the internship to the report and the defence

Every internship ends with a report and a defence. The report works best when it follows a simple logic: context, question, method, results, limits and personal lessons. Keep confidential material from the host organisation out of it: ask your supervisor what may be cited, and in what form.

The defence, for its part, tests your ability to explain your work to listeners who are not specialists. Prepare a presentation of a few clear slides with one main figure, and expect questions about what did not work. Outstanding research internships are sometimes recognised by the School at prize ceremonies, a sign that the quality of an internship is also measured by the rigour of how it is presented.