CazMeter: How CazVid Uses AI to Match Candidates with the Right Jobs
What CazMeter is and why it matters
CazMeter is the AI candidate matching engine built into the CazVid hiring platform. Every time a job seeker applies to a CazVid posting, CazMeter scores the fit between that candidate and the role on a clean 0-to-100 scale, surfaces the strongest profiles to the recruiter first, and explains the gaps in plain language. It is not a black box. It is a screening layer that takes the worst part of hiring, sorting through hundreds of applicants, and turns it into a ten-minute review.
For employers running high-volume pipelines, the difference is significant. A traditional resume stack is a list of files you have to open one by one. A CazMeter-ranked stack is an ordered queue with a percentage score, a video resume, and a one-line summary of where each candidate falls short. The same hiring decision that used to take days takes one focused sitting.
How CazMeter scores a candidate
CazMeter combines three signals into a single match percentage. None of the three is novel on its own. The leverage comes from running them together on the same applicant pool, in the same view, every time.
Skills and experience match. CazMeter compares the structured profile data on the candidate's CazVid profile (work history, skills, certifications, languages, location) against the structured fields on the job posting. The match score weights direct hits more heavily than partial overlaps. A candidate whose last role exactly matches the job title scores higher than one with three loosely related roles.
Pre-screening question answers. When the employer adds yes/no, short text, or short video pre-screening questions to a job, CazMeter folds the answers into the score. A "must have driver's license" yes/no with a "no" answer drops the candidate out of the top tier automatically. A short text answer about salary expectations or notice period gets parsed and compared against the job's preferences.
Video resume signal. The 30-second CazVid video resume gives CazMeter context that a written profile cannot. Communication style, language fluency, and presence are factored into the score using on-device transcription and lightweight signal extraction. The video is not used for any kind of biometric or demographic analysis. It is read for clarity, fluency in the role's required language, and whether the candidate's stated experience matches the recorded delivery.
The output is a single percentage between 0 and 100, plus a short text breakdown of which signal contributed which weight. Recruiters can sort by the percentage, filter by minimum score, or view the breakdown to understand why a specific candidate ranked where they did.
Gap analysis: the part recruiters did not know they wanted
The match percentage answers the question "is this person a fit?" Gap analysis answers the followup, "what would it take to make them a fit?" CazMeter generates a short gap analysis report for every candidate who lands above a configurable threshold but below a perfect match.
The report is two or three lines of plain language. Examples from real CazVid jobs: "Strong on customer service experience and Spanish fluency, missing two years of B2B sales experience." Or, "All required skills present, location 320 miles from job site, candidate has not indicated willingness to relocate." Or, "Skills match exactly, video resume in English at conversational level rather than native, role requires native English."
The report does three things at once. It saves the recruiter from reading every profile to figure out the same conclusion. It gives the hiring manager a structured way to compare candidates side by side. And it makes it obvious when a "missing" skill is actually a coachable gap rather than a hard disqualifier, which keeps strong candidates in the funnel who would otherwise have been auto-rejected by a keyword filter.
Pre-screening questions: text, yes/no, and short video
CazMeter is most powerful when the job posting includes pre-screening questions. CazVid supports three kinds:
- Yes/no questions. The simplest format. "Do you have a valid commercial driver's license?" or "Are you available for shifts on weekends?" Used to filter hard requirements.
- Short text questions. Open-ended answers up to a few sentences. "What is your expected starting salary?" or "Why are you interested in this role?" CazMeter parses the answers and folds them into the match score.
- Short video questions. Up to 120 seconds of recorded video answering a custom prompt. "Walk us through your most recent customer service success." Best for sales, customer-facing, and roles where communication is itself the skill.
Recruiters can require all three on the same job. The candidate sees them as part of the application flow on mobile, answers them in one sitting, and submits the application as a single bundle: 30-second video resume, structured profile, plus pre-screening answers. CazMeter scores everything together.
Most CazVid employers find that the right combination is two yes/no questions for hard requirements plus one short video question for the role-specific signal. That setup takes a candidate under three minutes to complete, which keeps the application drop-off rate low while still giving CazMeter the data it needs to rank well.
How CazMeter saves recruiters time
The honest answer is that CazMeter does not replace a recruiter. It changes which 10% of the work the recruiter actually does.
Without AI screening, a recruiter on a 200-applicant job spends most of their time on the bottom half of the funnel: opening profiles that turn out to be wrong, scheduling calls that turn out to be unfit, and writing rejection emails. The strongest candidates often get a slower response than the weakest ones because the recruiter is working through the queue in arrival order.
With CazMeter, the same recruiter opens an ordered list. The top 20 candidates by match percentage get reviewed in detail. The middle 60 get skimmed. The bottom 120 get a single batch rejection if none of them clear the minimum score. The recruiter spends their time where it matters: on the candidates most likely to convert, with the gap analysis already in front of them.
In practice, employers using CazMeter on the CazVid platform report screening 200 applicants down to 10 finalists in under an hour, compared to the multi-day process required to do the same on traditional ATS platforms. The time saved is not a vanity metric. It compresses time-to-hire, which directly affects the quality of the offer and the likelihood the strongest candidate is still available when the offer goes out.
The AI Job Builder companion
CazMeter only works as well as the job posting it scores against. A vague posting produces vague matches. A precise posting produces precise matches. Most recruiters do not have time to write a precise posting from scratch for every role, especially when they are filling a dozen openings at once.
The AI Job Builder is the companion feature that solves this. The recruiter types a few lines about the role, "senior customer success manager, remote, fluent Spanish, $90k base plus commission," and the AI Job Builder drafts a full posting: title, description, required skills, nice-to-have skills, suggested pre-screening questions, and a starter list of pre-screening yes/no filters. The recruiter edits the draft and publishes.
The output is structured the way CazMeter expects, which is the part that matters. A posting written by hand often misses fields that CazMeter uses to rank candidates. A posting from the AI Job Builder is complete by default. The end-to-end flow, from "I have a role to fill" to "the first 200 applicants are scored and ranked," takes less than 30 minutes for a recruiter with no AI experience.
What CazMeter does not do
It is worth being explicit about what is intentionally excluded from the scoring model.
CazMeter does not score on demographic attributes. The model does not use age, race, gender, religion, national origin, disability status, or any inferred proxy for those attributes. The video resume is read for fluency and clarity, not for facial features or demographic signals. CazVid runs internal audits to confirm the scoring distribution does not skew along demographic lines on any single posting.
CazMeter does not auto-reject candidates. Below-threshold candidates are deprioritized in the recruiter's view, but every candidate remains visible if the recruiter wants to scroll past the top tier. The recruiter makes the hire-or-reject decision. CazMeter ranks the queue.
CazMeter does not promise a perfect match. The percentage is a ranking signal, not a guarantee. A 92% candidate may turn out to be a wrong fit on softer criteria the model cannot read. A 68% candidate may turn out to be the right hire because of context the model does not see. Recruiters who treat the score as a starting point, not a verdict, get the best results.
Where CazMeter fits in the broader CazVid platform
CazMeter is one piece of a connected hiring stack. The candidate-facing piece is the 30-second video resume, which we cover in our 30-second video resume guide. The employer-facing piece is the desktop ATS, which uses CazMeter scores as the primary sort order on every job's applicant pipeline. New employers usually start with the high-level walkthrough in how CazVid works, then move to setting up their first posting through the dashboard at desktop.cazvid.app.
For recruiters concerned about platform legitimacy and data handling, our is CazVid legit guide covers the company, security posture, and what data is and is not used in scoring.
Pricing and how to try CazMeter
CazMeter is included on every paid CazVid employer plan, with the AI Job Builder and pre-screening questions bundled in. Free CazVid job postings include the basic match score; gap analysis reports and pre-screening question support are part of the paid tiers. There are no per-applicant costs, no per-score fees, and no separate add-on for the AI features.
The fastest way to see CazMeter in action is to post a job. Open desktop.cazvid.app, create an employer account, draft a posting with the AI Job Builder, and publish. The first applicants are usually scored within 24 to 48 hours, depending on the role and location. There is no demo gate and no sales call required to evaluate the product.
FAQ: CazMeter and AI candidate matching
What is CazMeter?
CazMeter is the AI candidate matching engine inside CazVid. It scores every applicant against every job on a 0 to 100 percentage and surfaces the strongest fits to the recruiter first, with a short gap analysis on the candidates close to but not at the top.
How accurate is the match percentage?
The match percentage is a ranking signal, not a verdict. It combines structured profile data, pre-screening question answers, and video resume signal into a single number. The intended use is to sort the queue, not to auto-hire or auto-reject. Recruiters who use CazMeter as a first-pass ranking tool typically find the top 20% of scores correlate strongly with the candidates they would have prioritized manually.
Does CazMeter replace recruiters?
No. CazMeter changes which work recruiters do. The recruiter still owns the decision on every hire. CazMeter handles the ordering, the gap analysis, and the bottom-of-funnel filtering, which frees the recruiter to spend time on the top candidates and the final interviews.
Does CazMeter use demographic data to score candidates?
No. CazMeter does not score on age, race, gender, religion, national origin, disability status, or any inferred proxy for those attributes. The video resume is read for fluency and clarity only. CazVid runs internal audits to confirm scoring distributions do not skew along demographic lines.
What is gap analysis and how is it different from a match score?
Gap analysis is a two- or three-line plain-language summary of why a candidate did not get a higher score. The match score is a number; the gap analysis explains the number. Together they give a recruiter both the ranking and the reasoning, which is the part that matters when comparing two candidates with similar scores.
Is CazMeter included in the free plan?
The basic match score is available on free CazVid job postings. Gap analysis reports, the AI Job Builder, and pre-screening questions are part of the paid employer plans. There are no per-score or per-applicant costs on any plan.
How long does it take to set up a CazMeter-ranked job posting?
Under 30 minutes for a first-time recruiter. Use the AI Job Builder to draft the posting, add two or three pre-screening questions, publish, and CazMeter starts scoring as soon as the first applications arrive. No technical setup or integration is required.
The bottom line
The bottleneck in modern hiring is not the lack of applicants. It is the time it takes to figure out which applicants matter. CazMeter exists to compress that step, surface the candidates most likely to convert, and explain the rest in plain language. It does not replace recruiter judgment. It replaces the busywork that gets in the way of recruiter judgment.
If you have an open role and want to see how CazMeter ranks your real applicants, post the job at desktop.cazvid.app. The first scores arrive within 24 to 48 hours, and the AI Job Builder gets the posting from "rough idea" to "live and ranked" in under an hour.