Get in touch
Home OverviewVideo Interview Features Trust All productsEight products, one data layer
Get in touch See how scoring works

Products › Video Interview

Every applicant gets an interview.
Not just the shortlist.

Structured video rounds, scored on one rubric

A recorded round every applicant can take the day they apply — same questions, same rubric, same scoring. So the shortlist is built from evidence rather than from whose CV happened to be read first.

Book a demo See how scoring works

Runs in the browser. Nothing for the candidate to install.

Conversational AI Avatar InterviewAsks the fortieth candidate exactly what it asked the first.Recorded Video InterviewThey record in their own time. You review in yours.2-Way InterviewYour own team, live, with the rubric on screen.

Three ways to run a video round, one rubric behind all of them — so an avatar score, a recorded score and a live score sit on the same axis.

A recording can be
re-watched. A phone screen cannot.

The first-round phone screen exists in one person’s memory and a few lines of notes. A recorded round is the same interview for everyone, reviewable by anyone on the panel, with the score attached to the moment that produced it.

  • The candidate answers on their time; you review on yours. Which is what makes interviewing every applicant possible instead of interviewing the ones you had slots for.
  • Every score links to its moment. A rating is a claim, and the claim points at the seconds of video that support it — so a panel disagreement is settled by watching rather than arguing.
  • Non-verbal signal, made structured. Posture, eye contact, gestures and composure are read and reported as reviewable intelligence rather than an interviewer’s gut feeling written down afterwards.
  • Hiring managers rate the round itself — not a recruiter’s summary of it. Weighted scorecards mean five people can assess the same candidate against the same criteria.
  • Two modes for the harder rounds. An avatar assessor runs the identical script for every candidate, and a two-way live round puts your own team in the conversation — both scored the same way as the async round.
Scorecard — senior backend engineer
Problem solving4.4
System design3.7
Communication4.6
Composure4.0
Ownership3.2
EvidenceSystem design 3.7 → jumps to 04:12
“I’d shard on tenant, but I haven’t had to”
214 applicants, all interviewed the same day · every score is a link, not an opinion

Applied this morning. Scored by lunch.

Five stages, no manual hand-offs. The invitation goes out on its own, and the scorecard is built from everything recorded along the way.

Invite sent
automatic
System check
1 min
Records
their time
Scored
one rubric
You decide

What the Video Interview does

The check that happens before the round+
Before a candidate records anything, they run a short check on camera, microphone and connection — and identity and document collection is handled in chat first. Two consequences: nobody loses a round to a broken headset, and you only ever review people who were validated and ready. The candidate does all of it in a browser, with nothing to install.
Question sets that adapt as they go+
Reusable question sets grouped into sections, with multiple-choice items mixed in where a knowledge check makes sense, and an interview that adapts based on what has been answered. Question generation runs in the background, so a newly opened role does not wait on somebody writing a bank of questions from scratch.
Scoring, and the evidence underneath it+
Every answer is scored against your rubric by a model tier reserved for scoring, and validated against a fixed schema rather than accepted as free text. Each score links back to the transcript excerpt and the moment of video it came from — so a rating can always be traced to something the candidate actually said, and a panel can go and watch it.
Body language and sentiment, read carefully+
Emotional tone and confidence — posture, eye contact, gestures, composure — are turned into structured, reviewable signal. This is the analysis only a video round can produce, and it is presented as one input among several rather than a verdict. A human still makes every call.
Avatar and two-way live+
An avatar assessor is a composed, bias-free interviewer that runs the identical script for every candidate, which removes the variance between a first interview and a fortieth. A two-way live round puts your own team in a real conversation. Both are scored against the same rubric as the async round, so a live score and a recorded score sit on the same axis.
Pipelines, leaderboards and centralised evaluation+
Candidates move through pipelines with weighted scorecards and leaderboards, so evaluation lives in one place instead of five inboxes. Hiring managers rate the same recorded round rather than a summary of it, and transcripts and analytics are retained for reporting on the process itself.

A recording of somebody’s
career

Video is the most sensitive thing you will hold about a person who does not yet work for you. Three constraints, all structural.

A human makes the call

It ranks and evidences. It does not reject anyone. The hire, the pass and the pipeline stage are decisions a named person takes, on the record.

One AI gateway, no loose ends

Every model call passes through a single choke point with schema validation, caching, per-tenant rate limits and spend caps. There is no second path to a provider.

Media behind signed links

Recordings sit in private storage reached only through short-lived signed URLs, with lifecycle rules that expire media on a schedule rather than keeping it indefinitely.

And plainly. A video round needs a working camera, a microphone and a connection — which is precisely why the system check runs first, and why the Voice and Chat rounds exist for candidates and roles where a camera is the wrong ask. We publish no accuracy figure for the scoring, because the honest measure is how often a human changed a score in your own pipeline — a number from your data rather than a claim about ours. Sentiment and body-language analysis is an input, not a verdict, and should never on its own be the reason somebody is passed over.

Voice InterviewThe same rubric without a camera — for roles where speech is the skill.Intelligent Recruitment SoftwareWhere the scorecard becomes a pipeline stage, and a stage becomes a signed offer.

Frequently asked questions

What does a candidate need to take a video interview, and do they have to install an app?+
Nothing to install: the candidate does all of it in a browser. They need a working camera, microphone and connection, and before recording anything they run a short check on all three, with identity and document collection handled in chat first, so nobody loses a round to a broken headset and you only review people who were validated and ready.
How are the answers in a video interview scored?+
Every answer is scored against your own rubric by a model tier reserved for scoring, and the output is validated against a fixed schema rather than accepted as free text. Each score links back to the transcript excerpt and the moment of video it came from, so a rating can always be traced to something the candidate actually said.
Does the video interview analyse body language, and can that decide the outcome on its own?+
It reads emotional tone and confidence (posture, eye contact, gestures and composure) and turns them into structured, reviewable signal. That signal is presented as one input among several rather than a verdict, and a human still makes every call.
Can I run a live two-way interview or an avatar-led interview as well as the recorded round?+
Yes. Besides the recorded async round there is an avatar assessor that runs the identical script for every candidate, and a two-way live round that puts your own team in a real conversation. Both are scored against the same rubric as the async round, so a live score and a recorded score sit on the same axis.
Where are candidate video recordings stored, and are they kept indefinitely?+
Recordings sit in private storage that is reached only through short-lived signed URLs. Lifecycle rules expire the media on a schedule rather than keeping it indefinitely.
When is a video interview the wrong choice, and how accurate is the scoring?+
For candidates and roles where a camera is the wrong ask, TalbotIQ's Voice and Chat rounds exist instead of the video round. TalbotIQ publishes no accuracy figure for the scoring; the measure it points to is how often a human changed a score in your own pipeline.

Published 9 September 2026 · Updated 9 September 2026 · Written by the TalbotIQ team

Ready to see it?

Tell us the role and your applicant volume, and we will run a full round against it — invite to scored shortlist.

hello@talbotiq.com
0 / 250
Request demo

Interview all of them.

Book a 30-minute demo and bring a shortlist you already made. We will show you what a full round would have surfaced instead.

Book a demo Talk to us

Malaysia-based team · Response within one business day · No obligation