Present and Defend Your Research, Then Explain Your Research Interests

This interview has no coding or algorithm question. The whole conversation is about your research, in three steps: 1. You introduce your own research. 2. The interviewer asks follow-up questions about the details of your project. 3. The interviewer asks which research directions you are interested in. Prepare both halves: an account of your research that holds up under detailed questioning, and an answer about your interests that connects to the work of the team you would join. ### Clarifying Questions - How long should the opening introduction run before the interviewer starts asking questions? - Should the introduction survey all of my projects, or go straight into one project in depth? - Is the interviewer an expert in my subfield, or a researcher from a neighboring area? - What problems does the team work on, so that I can relate my interests to them? ### Part 1 — Introduce your research and answer the deep-dive questions Introduce your research, then answer the interviewer's follow-up questions on the details of a project: why you chose the approach, how you know the results hold, and what exactly you did yourself. ```hint Problem before method Make sure the listener knows what problem you attacked, why it matters and what made it hard before you describe your method, and prepare one level of detail deeper than anything you say. ``` #### What This Part Should Cover - A problem statement and motivation that a researcher outside the subfield can follow - Your own contribution, separated clearly from the work of collaborators and advisors - The reasoning behind the key design and evaluation choices, including alternatives you rejected and the baselines you compared against - What the results do and do not show, including limitations and failed attempts ### Part 2 — Which research directions interest you The interviewer closes by asking which research directions you are interested in. ```hint Connect, do not list Names of popular fields say little. Think about which open question your own work left behind, and where it meets a problem the team works on. ``` #### What This Part Should Cover - Two or three specific directions, each grounded in your own work - An explicit link from at least one direction to the team's work - A sense of what could be achieved within an internship, and questions back to the interviewer about the team's priorities ### What a Strong Answer Covers - Depth calibrated to the listener, with the follow-up questions steering the conversation instead of a recited script - Claims about results tied to the setup, baseline and variance behind them - Clear ownership: what you decided, built and learned - Honest handling of questions you cannot answer - Interests that are specific, consistent with your record and connected to the team ### Follow-up Questions - If you had more time on this project, what would you do next, and why that before the alternatives? - Which result in the project are you least confident in, and what experiment would settle it? - What would have to change for your method to work on real production data or systems rather than a research benchmark?