Career services2 min read
Fresh graduates don’t need more interview advice. They need more interview practice
Career teams can’t run a mock interview for every student. AI voice mock interviews let graduates practise for real job adverts, with scores advisors see.
The pain point
“We have two career advisors and eight hundred final-year students. Most of them walk into their first interview never having practised out loud.”
Key takeaways
- Students rarely fail interviews for lack of ability; they fail because they have never answered the questions out loud.
- Career teams cannot offer one-to-one mock interviews at the scale students need.
- AI voice mock interviews let students practise as often as they like, for general interviews or for a specific real job advert.
- Advisors get scores and analysis across the cohort, so their one-to-one time goes to the students who need it most.
When we started talking to university and college career teams, I expected to hear about CV writing. Instead, I heard about silence: the pause after a graduate is asked “Tell me about a time you worked in a team,” and realises they have never said the answer out loud.
Interviewing is a skill, and skills come from practice. Advice helps. A tip sheet helps. But nothing replaces answering a real question, hearing yourself and trying again.
The career team’s maths problem
A one-to-one mock interview, with feedback, takes most of an hour of an advisor’s time. Multiply that by a final-year cohort and the numbers don’t work. So mock interviews go to the students who ask first, not necessarily those who need them most.
Meanwhile, the first round of a real hiring process is changing. Employers increasingly use structured, and sometimes AI-led, first-round interviews (I describe how ours works in the first-round phone screen). Graduates also use AI for their applications: Gartner’s research, reported by HR Dive, found that around four in ten candidates use AI during applications. A polished CV gets a graduate into the interview. Only practice gets them through it.
What good practice looks like
- Out loud, not on paper. Spoken answers expose gaps that written ones hide.
- Pitched at the right level. Questions for a fresh graduate should draw on studies, internships, part-time work and activities, not ten years of experience.
- Tied to real roles. Practising for the actual job advert a student is applying for is far more useful than generic questions.
- Repeated. The second and third attempts are where students improve most.
- Reviewed. Feedback is most useful when an advisor can see progress over time.
How Talenda School works
Talenda School gives a career team a student roster. Each student gets their own login to practise AI voice mock interviews pitched at fresh graduates.
- Personalised questions. Students keep their background and career goals up to date, and the questions are tailored to them.
- Practice for a real job advert. A student can add an advert they are preparing for. Talenda analyses its responsibilities, requirements and key skills, how it fits a fresh graduate, likely questions and preparation tips, then runs a mock interview for that role.
- Coach or realistic mode. Students choose between a coaching style of feedback and a more realistic interview.
- Scores and analysis for advisors. Each attempt receives an overall score and per-skill scores, with feedback. The career team sees results and AI progress analysis across the roster.
- Private notes stay private. Staff notes about a student are never sent to the AI.
Where the advisor’s time goes now
AI mock interviews don’t replace career advisors. They change what advisors spend their time on. Instead of running the same first-practice session hundreds of times, advisors can see which students haven’t practised, who is struggling with which skill, and who is ready. Their one-to-one time goes where it makes the biggest difference.
The best interview preparation has always been the same: practise, get feedback, try again. AI makes it possible to offer that to every student, not just the first in the queue.