How an AI Researcher Prepped Conference English in Moscow

Profile

A computer science major at Moscow State University, about 20, who had started AI research roughly five months before coaching. Her early results were on trustworthy and private models - defenses so modern systems do not leak sensitive data - with possible applications in healthcare, finance, and connected devices (IoT). A school-funded conference trip required an HR or researcher interview in English: clear motivation, a short research story, and the vocabulary to explain her work. Later classes shifted toward Academic IELTS and describing data as masters options came into view.

Pain Points

• She could explain her privacy and robustness research in short fragments, but not as a short English story that an interviewer could follow from start to finish.
• Her research introduction drifted into a vague “we” without saying whether she was speaking as an intern, a student, or both.
• She phrased results awkwardly - for example “quality lower just by 5%” - instead of a clear comparison such as “lower than the baseline by 5%.”
• Under pressure she mixed up common research English: trust to AI instead of trust in AI, how we can trust it instead of how can we trust it, reproducted instead of reproduced, and params instead of parameters.
• In long research discussions she overused “absolutely” and had little variety when agreeing or acknowledging a point.
• On the later IELTS path she needed work on v/w and wise pronunciation, and she treated advice as a countable noun.

Goals

• Build a strong 60–90 second introduction: role, research problem, approach, results, and next steps.
• Answer Why this conference? without sounding as if she only wanted to read papers at home.
• Explain privacy and robustness trade-offs and real-world applications in plain English.
• Handle follow-up questions: which methods come next, why this research area, and why this venue.
• Practise math and data English aloud: derivative, integral, fluctuate, anomaly, finite, and infinite.
• Later: build Academic IELTS Part 2 and Part 3 fluency for education and technology topics.

Lesson Notes

These are the classes we actually ran, not a generic curriculum. Each one started from a real interview question.

Feedback

Coach notes from those sessions, kept as they were given in class.

Common Mistakes

Corrections from live answers, the same patterns that showed up in class.

Before and After

The Result

These notes cover conference prep and later IELTS practice. No offer or acceptance is claimed. The result was a tighter research introduction, clearer motivation language for a funded trip interview, and stronger math and data vocabulary for academic English.

Advice for Early-Career Researchers Presenting in English

• If you are both a student and an intern, say that in sentence one - before you explain privacy trade-offs.
• Practise one clear baseline comparison aloud, such as “lower than the baseline by 5%,” until an interviewer can repeat it.
• Answer Why this conference? with networking and feedback, not “I can read the papers at home.”
• Rehearse the three follow-ups she prepared for: which methods come next, why privacy and robustness, and why this venue.
• Say derivative, integral, and parameters out loud. Conference question time will use them.

Why This Matters for AI Conference English

Conference and lab interviews punish vague ownership and soft metrics more than they punish accent. Non-native researchers who can narrate problem, approach, and result in calm English get the question time they need - whether the venue is an ICLR-style conference or a masters panel later.

Preparing for a similar path? Read Behavioral Interviews in English: A Backend Engineer Case Study.