Criterion: an AI interview coach that shows where every score came from
Practice an interview, see how each answer was scored, and challenge any score you disagree with.
Most interview practice tools give you a number and stop there. You can't see why you got it, and you can't disagree with it. Criterion links every score to the exact words you said that earned it. If you think a score is wrong, you can challenge it, and the model looks at it again in front of you. It can raise the score, hold it, lower it, or escalate it to a human.
An interview coach that links every score to the words that earned it, and lets you challenge it
Status
Live at criterionscore.com. New, and still early. Not tested with users
REC 02:14Product Designer
Interviewer
Walk me through a product decision you made where the data and your instinct disagreed. What did you do?
You
We saw retention climb after the redesign, but qualitative sessions told a different story. I ran a five person study before shipping wider, and it turned out the metric was masking a drop off I would have missed if I trusted the dashboard alone.
A time I changed my mind was...The tradeoff came down to...
Type your answer
Live readProvisional
Communication7.4
Product sense8.1
Craft & rigor6.8
7.4Tracking well reconfirmed at the end against the full transcript
Product roadmap
Static artifact
prototypes/roadmap.html
This is a static page shown in a frame. To use the real product, run the live app.
1Set up2Interview3Score4Debrief
The four steps a candidate goes through
The idea
Why I built Criterion
If a tool tells you that you scored a 6, you learn a number. You don't learn what you said, why it counted, or how to do better next time. So the job wasn't just to score an interview. It was to make the score something a person can check and push back on.
Five components do that work, and each one is shown below. They are the real shipped components, rendered from the live app's own code, not mockups. Next to most of them I explain the decision behind it: what I tried, what went wrong, and what I shipped. The sample content is the app's own demo interview.
Component 01 · Score Card, Confidence Meter
Score Card: each score with its confidence and weight
Most tools show a bare score like "6.5". Here each score also shows its status, how confident the model is, how much it counts toward the total, and why it landed where it did. The score most open to challenge gets a coral border, so you see it first.
Product Designer interview
Six questions answered · scored against your rubric
8.1
out of 10
Final
Strong across the board. Your product judgment carried the interview, and your reasoning was easy to follow. The one place to sharpen is rigor: back more of your claims with the concrete numbers you clearly have.
Weighted by what you set in setup: Communication 20, Product sense 30, Craft and rigor 30, Ownership 20.
Communication
20% of your score
8.4
Strong
Clear and well structured. You opened each answer with the tension, then walked the interviewer through your reasoning in order. Easy to follow without backtracking.
Question 3
You framed the conflict up front: the test said the winner was obvious, my instinct said we were about to ship a worse product. That gave every following point a clear place to land.
Question 4
You signposted your method plainly, I made myself prove it three ways before believing it, then delivered exactly three, in order.
Disagree? Tell the model what it missed.
Product sense
30% of your score
9.1
Strong
Excellent. You repeatedly separated the metric from the thing it was standing in for, and made decisions on the real outcome rather than the convenient number.
Question 3
You caught the proxy immediately: the test called success reached the dashboard, and you pointed out it said nothing about whether those people stuck around.
Question 1
You reframed the goal around value, not completion, it was never adoption, it was whether a conversation happened and came back.
Disagree? Tell the model what it missed.
Craft & rigor
30% of your score
6.8
Mixed
Solid method, lighter proof. You reached for controlled comparisons and downstream signals, which is exactly right. Several claims stayed directional, though, where you clearly had numbers to cite.
Question 4
Good instinct on isolation: I compared people who signed up through the new flow against people still on the old one in the same weeks.
Question 4
But the payoff stayed vague, the activated cohorts stayed separated for a full month, without the size of the gap or the sample behind it.
Disagree? Tell the model what it missed.
Ownership
20% of your score
8.0
Strong
Genuine. You named your own mistake without softening it, and framed the fix as something you would change in how you work, not just what the team should do.
Question 6
You owned the root cause directly: I let the test define success before we ever launched it, and that is what nearly shipped the wrong flow.
Question 5
You held yourself to a standard for next time, I would have written the rollback line before shipping, rather than blaming the process.
Disagree? Tell the model what it missed.
The real shipped component, rendered from the live app's code.
Component 02 · Evidence Drawer
Evidence Drawer: the exact words behind each score
Open any score to see the words that earned it. Each piece of evidence has a quote, why it moved the score, and a timestamp that replays the moment you said it. Whether it helped or hurt is shown with an icon and a written note, not by color alone. Hover a quote and the same words highlight in the transcript below.
Craft & rigor
30% of your score
6.8
Mixed
Solid method, lighter proof. You reached for controlled comparisons and downstream signals, which is exactly right. Several claims stayed directional, though, where you clearly had numbers to cite.
Question 4
Good instinct on isolation: I compared people who signed up through the new flow against people still on the old one in the same weeks.
Question 4
But the payoff stayed vague, the activated cohorts stayed separated for a full month, without the size of the gap or the sample behind it.
Disagree? Tell the model what it missed.
The real shipped component, rendered from the live app's code.
The decision: replay the words, not a video
TriedReplaying a video clip
The prototype played back a recorded clip of the candidate at the moment a score was set. It looked good in a demo, but it was wrong. Your camera view stays on your device and is never sent anywhere, and the product's promise is that you're judged on what you said.
BrokeIt went against the product's promise
Showing a face on replay brought back the thing the product is meant to remove. If the score is about words, the evidence has to be words.
ShippedThe quote and its context
Replay shows the exact quote a score is based on, the question it answered, when it was said, and what it matched. The replay says plainly what it is showing. If a real clip is ever recorded, it can go in the same spot. Until then, there's nothing to fake.
Component 03 · Challenge Panel, Challenge Thread
Challenge Panel: disagreeing with a score and getting it re-checked
If you disagree with a score, you explain why, and the model re-evaluates it right there. There are four possible outcomes: raise, hold, lower, or escalate to a human. Every challenge and its result stay on the page, so nothing changes out of sight.
What did the score miss?
Point to the moment, or pick a reason. The model re reads that part of the transcript.
The real shipped component, rendered from the live app's code.
The decision: make disagreeing a real option
TriedA feedback form
The obvious version was a "report this score" link that sent a complaint nowhere. People could vent, but nothing changed. That's worse than having no option at all.
BrokeA challenge has to be able to change the score
If a challenged score can never move, the challenge is just for show. The model has to re-evaluate live and be willing to change the score on the record.
ShippedA limited, visible re-evaluation
Each challenge re-scores against the current value, shows its reasoning, and is added to a visible history. It can raise, hold, lower, or hand off to a human. Challenges are capped, and the handoff is always there. Those two limits keep it from turning into haggling.
Component 04 · Interview Stage, Live Read
Interview stage: a live read that updates as you answer
You don't have to wait until the end to see how you're doing. You answer, and a live read updates each time you submit, using the same criteria as the final Score Card. It's marked Provisional because it gets rechecked at the end against the full transcript. The whole conversation stays on screen, so you can always see what's being scored.
Product Designer interview
Question 3 of 6
Walk me through a moment the data and your instinct disagreed. What did you do?
Your answerYour turn
0 / 1500
The live read updates when you submit.
Live read Provisional
Communication7.4
Product sense8.1
Craft & rigor6.8
7.4
Tracking well
reconfirmed at the end against the full transcript
Interview feed
Interviewer
What was the hardest tradeoff in that project, and how did you make the call?
You
Speed against clarity. I cut a personalization step that tested well because it added a screen most people skipped.
Interviewer
To start, tell me about a project you are proud of and your specific role in it.
You
I led the onboarding redesign. I owned it end to end, from framing the drop off problem to shipping the React flow.
The real shipped component, rendered from the live app's code.
Component 05 · Debrief Timeline
Debrief timeline: each answer linked to the scores it moved
The Score Card shows what each criterion earned. The debrief shows what each of your answers did to the scores, which is more useful. Every answer has chips (small tags) for the criteria it affected, marked as helped or hurt, and each chip jumps to that score. It's built from the same evidence as the Score Card, so the two can't disagree.
The coaching read
Product Designer interview
You interviewed like someone who trusts data but refuses to be fooled by it. Your product judgment set the ceiling, and your reasoning was easy to follow the whole way through. The distance between this and a great interview is proof: you reach for the right method, then stop just short of the number that would seal it.
In answer after answer, you asked what a number was actually standing in for before you acted on it. That instinct is the core of product sense, and it was the strongest thread of the interview.
The momentThe test called success reached the dashboard, and you pointed out it said nothing about whether those people stuck around.
You frame the tension before the answer
You opened with the conflict, which gave the interviewer a place to put everything that followed. It is why you stayed easy to follow even on the harder questions.
The momentThe test said the winner was obvious. My instinct said we were about to ship a worse product.
Put a number on the claims you already have
Your method was sound, but the payoff often stayed directional. You said a difference held for a month without ever saying how big it was or how many people it covered.
Next time: attach one number to each claim. The lift, the sample, the delta. It turns a good story into evidence.
The momentThe activated cohorts stayed separated for a full month, with no size behind the gap.
Close your answers
A couple of strong answers trailed off after the middle. The last sentence is the one the interviewer carries into the next question, and yours sometimes faded rather than landed.
Next time: end on the decision or the result, in one clean line. Say what you did and what it moved.
The real shipped component, rendered from the live app's code.
The decision: one set of evidence, read from both sides
TriedA separate debrief summary
Early on, the debrief was its own write-up, written after the scoring. It read well, but it could drift from the scores. It could say an answer mattered when the scores didn't show that.
BrokeTwo views that could contradict each other
When the debrief and the Score Card are generated separately, they can disagree. Then you can't trust either one.
ShippedOne source for both views
The debrief uses the same evidence graph as the Score Card (the saved links between each answer and the scores it affected), read from the transcript side. Each answer's chips are generated from that data, not written by hand, so they can only point to scores the cards already show.
Measurement plan
How I will know it is working
Criterion is live and early, so I don't have results to share yet, and I won't make any up. Instead, here are the four signals I'm watching in real use. I wrote this plan before any data came in, because choosing what to measure is a design decision too.
Legibility
Do people open the evidence?
If nobody reads the words behind a score, showing them doesn't help. I want to see people open the Evidence Drawer on a real share of scores, not just the low ones.
Accountability
Do challenged scores change for good reasons?
Challenging a score only builds trust if it sometimes changes the score for a good reason. I'm tracking how often scores are raised, held, lowered, or escalated, and reading the arguments behind each.
Completion
Do people reach the debrief?
The debrief is where answers connect to scores, and that's where most of the value is. If people stop at the number, they miss the most useful part.
Trust
Do people act on the result?
The real test is whether someone trusts the result enough to change how they prepare, or to save the Score Card and share it. That's what the whole design is meant to earn.
v1 to v12
What changed from v1 to v12
Each version made the scoring easier to read and easier to challenge. I removed things about as often as I added them. Each step below relates to a component shown above.
early
A bare number. A score and a one-line reason. Clean, but you couldn't learn much from it. Everything after this was built to fix that.
mid
Evidence added. Scores started linking to the exact words. The Evidence Drawer and the transcript highlight came in together. Replay was still a video clip at this point. I cut it later.
mid
Challenges added. The Challenge Panel turned disagreeing into a live re-evaluation with a visible history. The cap and the human escalation were added at this point.
now
No video, one source of evidence. The interviewer became an orb instead of a face, the video replay became a transcript quote, and the debrief was rebuilt on the same evidence as the Score Card.
Three more case studies about systems that explain themselves: a dashboard broken down one
design choice at a time, an incident console that finds the cause, and a compliance agent that
never hides a failure.
A data story about the Kobe, Tohoku, Kumamoto and Noto earthquakes, with live charts built on US Geological Survey data. Every chart opens for a closer look, and the dashboard behind them is public.
When one fault sets off hundreds of alerts, Vantage traces them back to the one that started it, shows the path it took through the system, and points to the slowest step.
Scans a website against the European accessibility standards (EN 301 549 and WCAG), opens tested code fixes as pull requests, and records every step. It never says a site is compliant, because only a legal review can say that.