Research internship and Master's thesis interview preparation

Why it matters

For a Master's thesis or research internship, the conversation decides not only acceptance but also which project and supervisor you get. Unlike many employment processes, academic labs do not share a standardized interview loop: one professor holds an informal chat, another asks for a short presentation of past work, another sends a paper to discuss or a small task, and some add a coding exercise. Company research internships vary as well and often include coding interviews similar to software-engineering loops alongside research or project discussions. Ask the lab or recruiter what to expect, and prepare the common core below.

Core ideas

What they want to learn. Can you think about research questions concretely? Do you know the area well enough to start? Can you execute (code, run experiments, read papers)? Will you work reliably at the expected level of independence? Do your interests and timing fit an available project?

Preparation checklist.

1. Research interests: three sentences naming an area, a specific question and why it matters to you. Specific beats broad ("tail latency under co-location in serverless platforms" rather than "cloud").
2. A possible problem: one proposal shaped by the Heilmeier questions, with a first experiment you could run in two weeks, the main risk and a fallback. Hold it loosely; supervisors often steer you to an existing project.
3. Motivation for this lab: read its two or three most recent relevant papers and project pages; know its systems, datasets or testbeds. Explain why this group rather than the area in general.
4. Relevant papers: discuss two papers in depth from claim-evidence notes, including one weakness and the experiment that would address it.
5. Feasibility: what fits the thesis or internship period (check your program's rules on duration and credits), the data, compute, cluster or hardware access required, and skills to learn first.
6. Supervision expectations: meeting frequency, day-to-day mentor (often a PhD student or postdoc), expected independence, whether publication is a goal, industry co-supervision and how the topic gets fixed.
7. Prior technical work: two- and ten-minute explanations of your strongest projects and coursework, with your exact role, one decision you would revisit and what you measured. Use only real details.
8. Logistics: start date, duration, location, funding or salary and any visa questions.

Questions you may be asked.

• Why this lab, and which of our papers relates to your interest?
• Walk me through your most technical project. What was hard, and what would you change?
• Describe a paper you read recently. What is its main weakness?
• How would you evaluate this idea, and against which baseline?
• What would you do if the main approach fails after a month?
• Which tools and languages are you fluent in? When can you start, and how much time can you commit?
• For company internships: coding problems, and possibly a design discussion depending on the team.

Questions you should ask.

• What does a successful thesis or internship look like here, and can I see a past example?
• How are topics chosen, and how much can they change?
• Who supervises day to day, and how often do we meet?
• What infrastructure and compute are available, with what constraints?
• Are students expected to publish, and how is authorship decided?

Company research internships. Read the official program pages (for example the Google Research and Microsoft Research careers pages) for role types and eligibility, and confirm with the recruiter whether interviews include coding, research discussion, a presentation or team matching. Keep the coding lane of your plan active if they do.

Common misconceptions

• "There is a standard thesis interview to drill." There is not; confirm the format and prepare the core.
• "I must arrive with a finished thesis proposal." A concrete, revisable idea plus evidence that you can execute is usually worth more.
• "Praising the lab shows motivation." Specific knowledge of its recent work does.
• "Research internships do not test coding." Many company programs do; check rather than assume.
• "Overstating past results helps." Supervisors probe details; precise, modest accounts build trust.

How to approach it

Use the app's Thesis prep page for this checklist, Research notes for paper notes and Project notes for explanations of prior work. Rehearse with a peer playing the supervisor using both question banks, then refine the proposal after reading the lab's latest papers.