Manual review load
Reviewers repeat the same questions across hundreds of candidates and lose time before real shortlist work begins.
NewAI voice interviews for selection teams
AI interview screening
Run first-round interviews without running first-round interviews.
Configure a reusable interview agent, share a booking link, candidates complete a live voice + video interview, and you receive trait-based reports with quotes and flags.
Start pilot. No credit card required.
Report in ≤ 2 minutesQuotes as evidenceVoice + video recordedNo credit card required
Report · 1m 47s
Score87JT
Jordan Torres
Technical Fellowship · Round 1
Communication
88
Problem-solving
82
Initiative
91
“I led the rollout across three markets without a dedicated PM in the loop...”
Quote evidenceVerify depth
Strengths · Weaknesses · Follow-upsCompleted
Example from a recent selection round
0
Applicants
then
0+
Videos reviewed
then
0
Calls advanced
Trusted by selection teams
RiseSelection TeamsAdmissions OpsFellowship ReviewBootcamp IntakeTalent AcceleratorsScholarship CommitteesProgram DirectorsAccelerator OpsRiseSelection TeamsAdmissions OpsFellowship ReviewBootcamp IntakeTalent AcceleratorsScholarship CommitteesProgram DirectorsAccelerator Ops
The problem
Teams lose hours of manual, exhausting work to scheduling overhead, review backlog, and inconsistent first-round screening.
Scheduling friction
Calendars bottleneck the pipeline and force interview operations to revolve around recruiter availability.
Inconsistent scoring
Without a shared rubric and evidence trace, first-round decisions are hard to compare and hard to defend.
How it works
Configure. Share. Review.
Configure
Set program context, ideal candidate or job profile, traits, weighted rubric, questions, and report structure.
Operator-defined setup
Share
Create one booking link with expiration, intake questions, and access code. Candidates see recording disclosure before start.
Link-first distribution
Review
Open strengths, weaknesses, follow-ups, and quotes mapped to traits in a report delivered in ≤ 2 minutes.
Decision-ready output
Exact configuration order
Context→
Ideal candidate→
Traits→
Rubric weights→
Questions→
Report structure
Product
It's Calendly for AI interviews—built for selection teams.
What operators see
Before the session
Booking link with access code and recording disclosure. Candidates confirm terms before interview start.
Report delivery
Trait-based report in ≤ 2 minutes. Strengths, weaknesses, follow-ups, and quotes as evidence.
Dashboard state
Live session tracking: In Progress, Completed, Failed. No refresh loops.
This is not an ATS.
Reusable Interview Agents
Persona and evaluator behavior are configured once at the agent level and reused across programs.
Versioning keeps rubric changes auditable.
Learn moreCalendly-style booking
Generate one link for candidate scheduling with custom windows, expiry, and access code controls.
Your team still owns outreach channels.
Learn moreLive voice + video
Candidates complete a live session. Voice + video are recorded with disclosure before interview start.
Video is used for proxy protection, not facial recognition.
Learn moreTrait-based scoring
Scoring follows your operator-defined traits and rubric weights, not one-size-fits-all defaults.
Accent and fluency are not penalized by default.
Learn moreReport in ≤ 2 minutes
Each completed interview produces strengths, weaknesses, and follow-up prompts quickly.
Quotes as evidence are linked to every key claim.
Learn moreOperator dashboard
Track session states in real time: In Progress, Completed, Failed.
This is not an ATS. Pruun is a focused screening layer.
Learn moreWhy it works
Signal
Candidate authenticity
Access codes plus live voice + video reduce proxy interview risk.
Operator-controlled accessOps
Capacity without calendar drag
Run first rounds continuously without creating more recruiter call blocks.
Async-first, no scheduling overheadEvidence
Explainable outcomes
Every report includes quotes as evidence, with trait-level score rationale.
Linked to transcriptControl
Calibration loop
Human reviewers sample outcomes and tighten rubric behavior over time.
Human-in-the-loop by designAccelerators
Move from thousands of applicants to shortlist faster while preserving evidence quality.
Bootcamps
Apply one interview bar across intakes and reviewers in different time zones.
Fellowships
Defend selection decisions with reproducible trait-based reporting.
Operator feedback
“We stopped burning nights on first-round calls. Reports arrive with quotes as evidence, and shortlist meetings are now concise and defensible.”
Program Director, technical admissions team
Trust
Operator-defined rubrics. Calibration loop. Human-in-the-loop review.
Scoring philosophy
Operator-defined rubrics, weighted traits, and a calibration loop with human-in-the-loop review.
Bias stance
Accent and fluency are not penalized by default. Model and prompt versions are tracked for reproducibility.
Explainability note
Reports map claims to quotes as evidence, so reviewers can trace why a candidate advanced.
Voice + video are recorded with clear disclosure before interview start. Video is used for proxy protection, not facial recognition. Most reports arrive in ≤ 2 minutes with quotes as evidence.
Pricing
Simple, usage-based pricing. Pay per completed interview.
Credit packs
Usage-based credits billed per completed interview.
Program bundles
Pre-sized bundles for specific intake windows and hiring cycles.
White-label premium
Custom voice and tone plus branded candidate link and interview UI.
Plan your next intake with Pruun.
Share expected applicant volume and timeline. We will recommend the right credit structure.
FAQ
Common rollout questions
No. Pruun is a screening system for first-round interviews and reporting, not an ATS replacement. It sits upstream of your ATS and feeds shortlist decisions into whatever workflow you already use.
No. Pruun is link-first. We generate the booking flow and your team distributes links in your existing channels—email, portal, Slack, or wherever applicants already look.
Yes. Voice + video are recorded with disclosure before interview start. Video is for proxy protection, not facial recognition. Candidates are informed before the session begins.
Most reports are generated in ≤ 2 minutes and include quotes as evidence. You see strengths, weaknesses, follow-up prompts, and scores mapped to your rubric.
Scoring is driven by your operator-defined rubric and weighted traits. Accent and fluency are not penalized by default. Model and prompt versions are tracked, and a human calibration loop lets you tune behavior over time.
Single next step