Skip to content
For university career services

Interview preparation that moves as fast as your students' career goals.

Talvin gives every student unlimited practice interviews with detailed, specific feedback — so they build confidence before it counts, without overwhelming your career services team.

Unlimited practice, any hour · Institution-wide licensing
Practice session · in progress
23:12 · nobody watching
Final year · practice attempt 3
Engineering · graduate scheme
Walk me through a system you designed where you had to trade off cost against reliability.
We had a sensor rig that kept dropping readings, so the team rebuilt the buffer…
What did you personally decide there, rather than the team?
feedback, the moment it ends
Answer structureneeds work
Domain knowledgestrong
Clarity out loudimproving

Career services are outnumbered.

Delivering personalised coaching to thousands of students with a team of a few is not a scheduling problem. It is an arithmetic one.

70%
of students feel unprepared for their first industry interview, because they have never practised one.
Never practisedHave
10+ hours
a week of advisor time spent running basic mock interviews, leaving less for the career strategy work only a human can do.
One advisor week
Repeat mock interviewsEverything else
“Be more
confident”
is the feedback most students actually receive. It doesn't tell them which answer was weak or what to do differently.
Description of a symptom
“You seemed nervous. Be more confident next time.”
Instruction they can act on
“Answer 3 described what the team did. Say what you decided.”
The students who need practice most

Appointments favour the students who already know how to ask.

Every career service runs on the same quiet filter. The students who book mock interviews are the confident ones, the well-networked ones, the ones who already understand that practice is what makes the difference.

The students who most need the practice — first in family to attend university, international students interviewing in a second language, anyone who suspects they will be bad at it and would rather not find out in front of a member of staff — are the least likely to put their name down.

Access stops depending on availability
Who gets through the booking gate
Confident, well-networkedbooks
Already interview-readybooks
First in familyrarely
Second-language speakerrarely
Expects to be bad at itno
Need runs the opposite way to uptake. The gate is not the calendar — it is having to admit out loud that you need the help.
With the gate removed
Everyone, at 11pmall

Practice that feels like the real thing.

Spoken, role-specific and private — the three things a mock interview needs to be useful, and the three hardest to deliver at scale.

Spoken, not typed
No edit buttonlive
Follow-ups adaptyes
Realistic voice conversations
Natural, spoken interviews that mirror the pressure and pacing of a real recruiter — not a form, and not a list of questions to read.
Pick a discipline
Engineering · graduate scheme
Walk me through a system you designed where you had to trade off cost against reliability.
Technical depthDesign trade-offsOwnership language
Industry-specific scenarios
From engineering to nursing, business to humanities, role-specific questions that test actual domain knowledge rather than generic competencies.
Three attempts, one evening
21:4038
22:2561
23:1279
PrivateUnlimited retriesNo staff present
A safe space to fail
Students can stutter, pause, lose their thread and start again in private. Failing here is how they stop failing when it counts.
CommunicationAnswer structureDomain knowledgeClarity
Every session ends with a report analysing all four — with specific things to change, not encouragement.
What your team gets

Capacity back, and a clearer picture of who needs you.

Unlimited practice without unlimited staff.

Whether you serve 500 students or 50,000, practice capacity is the same. No scheduling bottleneck, no waiting list, no additional headcount.

One 12-week term12,000 students
By appointment · 4 advisors, 8 mocks a week each384 sessions
Reaches 3% of the student body, once each.
With Talvin · same teamno ceiling
Every student, as many attempts as they want.
The gap does not close by working harder. It closes by removing the appointment.
An advisor's week, before and aftersame hours
Now
Repeat mocks
Admin
Strategy
Employers
With Talvin
Triage
Admin
Career strategy
Employer advocacy
The repetitive layer moves. The relationship stays.
Time for the work only humans do.

Career strategy, industry connections, difficult conversations about direction, advocacy with employers. Repetitive mock interviews stop consuming the hours that work needs.

Visibility into who is struggling.

See which students are practising, where they are improving, and who is repeatedly stuck — so your team's limited human time goes to the students it will change the most.

Practising, but not improvingflag for a human
M. Osei · BEng · 2 sessions improving
A. Kaur · BSc Nursing · 6 sessions stuck
J. Lindqvist · BA History · 4 sessions ready
A more accurate way to allocate advisor time than first-come-first-served booking.
Readiness by programme · ahead of recruitment season
Engineering81%
Business74%
Nursing & allied health62%
Humanities48% · concentrate here
Comparable across cohorts, programmes and years — the same question you are asked in every quality review.
Evidence of readiness.

Structured, comparable data on interview preparedness across cohorts, programmes and years.

Set it up once. Students use it whenever they need it.

Configuration is a one-off job for your team. After that, the practice happens without you in the room.

01Configure it for your programmes

Set up interview scenarios by discipline, industry and level — graduate schemes, placements, internships, professional roles. Start from templates and adjust to what your students are actually applying for.

Scenario builder
DisciplineMechanical Engineering
LevelGraduate scheme
FormatTechnical + competency
Employer templateLoaded
02Give students access

Students log in and practise on their own schedule, as often as they want. Nothing to book, nothing to install, no waiting for an appointment.

When practice actually happens
09:00office hours23:00
Most sessions start after your team has gone home. None of them needed a slot.
03They get specific feedback immediately

Every session produces a report on communication, structure, domain knowledge and clarity — with concrete improvements rather than general encouragement.

Session reportready in seconds
72
Comms
54
Structure
88
Domain
69
Clarity
Q3: you described the outcome but not your method. Name the two options you rejected.
Full transcript attached — read back what you actually said.
04You see the picture across the cohort

Track participation, improvement over time, and which students may need direct support from your team.

Participation · final-year cohortweek 1 → 8
Each square = 25 studentspractised at least twice
Outcomes you can report on

Employability is measured. This is measurable.

Graduate outcomes are not a soft metric for a university. They feed rankings, they appear in accreditation reviews, and they are increasingly what prospective students compare when choosing where to study.

Interview performance is one of the few points in that chain a career service can directly influence — and until now, one of the hardest to scale or evidence.

The graduate-outcomes chain
One link in the chain sits inside your control. It has also been the one you could never evidence at scale.

From engineering to humanities.

Talvin adapts to the industry and role a student is actually pursuing. Students preparing for a specific employer can practise against that role rather than a generic template.

Technical
Engineering & computer science
Design trade-offs, debugging reasoning, and explaining a system out loud under questioning.
Technical deep-diveGraduate scheme panelPlacement screen
Clinical
Nursing & allied health
Scenario questions on patient safety, escalation and working inside a hierarchy.
Placement interviewValues-based recruitmentClinical scenario
Case-style
Business & consulting
Structuring a problem live, doing arithmetic out loud, and defending a recommendation.
Market-sizing caseProfitability caseInsight / IB first round
Competency
Public sector, education, humanities
Structured competency questions and the graduate schemes most humanities students apply into.
Civil service frameworkTeacher trainingCharity & NGO grad scheme
Practising for one specific employer? Point the scenario at that role — the format, the competencies and the questions that employer is known to ask.
Practising for Grad scheme · assessment centre round 1 Configured

Built for how universities actually buy.

Licensing, data protection and procurement — the three things an institutional evaluation turns on.

Scope of access
Whole institutionmost common
A facultyoptional
A programmeoptional
One cohort · pilotstart here
Institution-wide licensing
Access for your entire student body, or scoped to specific faculties, programmes or cohorts.
Encrypted in transitTLS
Encrypted at restAES
Never used to train modelsNever
Retention windowYour policy
Student access to own historyConfigurable
Built around GDPR and the data-protection frameworks of the markets we operate in.
Data protection
Student data is encrypted in transit and at rest, and is never used to train our models. Retention is configurable to your institution's policy.
Evaluation pack
Security documentation
DPIA support
Accessibility statement
Tender responses
Invited to bid on international university tenders.
Procurement and tender
We work through institutional procurement processes. If you need documentation for a formal evaluation, we can provide it.
FAQ

Questions from career services teams

The ones that come up on almost every university call.

How is this different from a student practising with a friend, or with ChatGPT?

Practising with a friend is genuinely useful, and students who have someone willing to do it should keep doing it. The problem is that most students do not — and the ones who do tend to be the ones who least need the help.

A friend also cannot give structured feedback. They can tell a student they seemed nervous; they cannot tell them that their answer to the third question described what the team did rather than what they personally contributed, which is the specific habit costing them offers.

Text-based AI tools have a different limitation. A student typing an answer has time to think, edit and rephrase — none of which is available in a real interview. The difficulty of an interview is not composing a good answer; it is composing one out loud, immediately, while someone waits. Voice practice rehearses the thing that is actually hard.

Talvin also adapts. When a student gives a vague answer, the follow-up presses on the vague part — the same thing a real interviewer does, and the moment most students are least prepared for.

Do students actually use it?

Usage tends to be highest in the weeks before recruitment season and around application deadlines, which is when it matters most.

The pattern career teams describe is that removing the booking step changes who participates. An appointment requires a student to decide in advance that they need help, find a slot, and turn up — three points at which a nervous student drops out. On-demand practice at 11pm requires none of that.

It also helps that students can repeat it. Very few will book three mock interviews with a member of staff. Many will do three practice sessions in a week when nobody is watching and there is no cost to doing badly.

Framing matters. Uptake is significantly higher when it is introduced as part of a programme or module rather than listed as one more optional service on a careers portal.

What kind of feedback do students actually get?

A report after every session covering how they communicated, how they structured their answers, the domain knowledge they demonstrated, and how clearly they made their points — with specific things to change.

The distinction that matters is between description and instruction. “Be more confident” is a description of a symptom. Useful feedback identifies which answer was weak, what was missing from it, and what a stronger version would have included.

Because the whole session is transcribed, students can also see exactly what they said. Most are surprised — people rarely remember their own interview answers accurately, and reading them back is often the moment the feedback lands.

Can we tailor it to our programmes and the employers our students apply to?

Yes. Scenarios are configured by discipline, industry and level, and you can build them around the employers and schemes your students actually target.

The versions that work best are usually specific: the graduate scheme that takes twenty of your students a year, the clinical placement interview with a known format, the technical screen a particular employer is known to run. Those give students rehearsal against the real thing rather than a generic approximation.

If you have employer partners willing to describe their interview format, that is worth building into the scenarios directly.

How does this fit with the coaching we already do?

It replaces the repetitive layer, not the relationship.

The typical pattern is that Talvin handles volume practice — the reps a student needs to stop being bad at interviews — while your team keeps the work that requires judgment: career direction, difficult conversations, employer advocacy, and supporting the students who need more than practice.

Several teams use the data as a triage layer. Students who have practised repeatedly without improving are surfaced for a human appointment, which is a more accurate way of allocating scarce advisor time than first-come-first-served booking.

What can we see as an institution?

Participation, improvement over time, and readiness levels across programmes and cohorts.

That answers questions career services are often asked and rarely have data for: which faculties are engaging with preparation, whether readiness is improving year on year, and where to concentrate resource ahead of a recruitment cycle.

It is also reportable evidence for accreditation reviews and internal quality processes — an area where career services provision is frequently strong in practice and weak on documentation.

Is student data safe?

Student data is encrypted in transit and at rest, and is never used to train our models.

Retention is configurable to your institution's policy rather than ours, so recordings and transcripts persist only as long as you require. Students can be given access to their own session history.

Talvin is built around GDPR and the data-protection frameworks of the markets we operate in. For institutions with specific obligations — sector regulators, research ethics requirements, or national frameworks — we can work through those as part of procurement.

How is it licensed for a university?

Institution-wide access is the most common arrangement, though licensing can be scoped to particular faculties, programmes or cohorts if you want to start narrower.

For universities, pricing is handled as an institutional agreement rather than through the standard published plans, because the usage pattern is different — heavily seasonal, spread across a large population, and tied to an academic year rather than a hiring cycle.

We work through institutional procurement processes and can provide documentation for a formal evaluation. Book a demo and we can size it against your student numbers.

Can we run a pilot with one cohort first?

Yes, and it is usually the sensible way in.

The most useful pilots are tied to a specific cohort with a recruitment deadline — a final-year group heading into graduate applications, or a placement year cohort. That gives a clear before-and-after within one term rather than a general trial with no natural end point.

Run it alongside your existing provision rather than replacing anything, and compare participation against what your team can currently deliver by appointment.

Start with one cohort.

Pick the group heading into recruitment season, give them unlimited practice, and see how many take it up compared with how many book an appointment today.

Institutional agreement · scoped to one cohort or the whole student body