Project Brief — Full DetailEngagement

AI Interviews. Human Decisions.

Series A · AI Interview Platform · Live & Pilot-Ready · Confidential · Jan 2026

Voice/video AI interviewers hold adaptive conversations with candidates; a scoring engine returns evidence-backed assessments; an integrity layer flags cheating for human review. Humans always decide — Mentis assesses and recommends.

Platform Capabilities

Assessment

Adaptive Interviews

AI voice/video interviews (Lisa & Bob personas), GPT-4 adaptive questioning, 24/7 self-serve.

Intelligence

Evidence-Backed Scoring

Evidence-backed scoring across skill, communication & fit; cheat detection at 95%+ accuracy.

Foundation

Human-in-the-Loop

Human-in-the-loop decisions, SOC 2 / GDPR / HIPAA-ready, Greenhouse ATS sync.

Technical Architecture

Job DescriptionRecruiter upload

CandidateSelf-serve link

ATSGreenhouse / Lever

Mentis Assessment EngineBlueprint → Interview → Scoring

Evidence ScorecardRecruiter decision

ATS Sync + AnalyticsWrite-back & dashboards

1 · CREATE · 30 SEC

Create

Upload a job description; AI generates a structured interview blueprint.

2 · INTERVIEW · 15–30 MIN

Interview

Candidates interview 24/7; AI adapts follow-ups in real time.

3 · DECIDE · 5 MIN

Decide

Recruiters review evidence-backed scorecards to make the call.

WK 1–2

Launch Prep

Video AI stability, public launch

WK 3–4

Initial Adoption

First signups, onboarding tuning

WK 5–6

Acceleration

Multi-language, +20 templates

WK 7–8

Optimization

In-app onboarding, Enterprise pilot

WK 9–12

Quarter Exit

2,000+ interviews, NPS 60+

Competitive Position

vs. HireVue (~$700M)

AI-native voice + video at ~1/10th the price.

vs. Greenhouse (~$4B)

The interview-intelligence layer the ATS lacks.

vs. AI Wrappers

Enterprise infrastructure & compliance, not a weekend wrapper.

Security, Privacy & Trust

Tenant Isolation

Row-level security — no cross-tenant read path.

Consent-First

Explicit consent to recording & AI assessment.

Encrypted End-to-End

TLS 1.3 in transit, AES-256 at rest.

Compliance-Ready

SOC 2 / GDPR-aligned, HIPAA-ready.

The Challenge

20+ hours per hire wasted on screening; 60% of candidates fail basic screens; small teams can't match Big Tech recruiting scale; human interviews stay inconsistent and biased.

Our Approach

A 24/7 AI interviewer applying an identical rubric to every candidate, with evidence-linked scores and advisory integrity flags — humans make every final call.

Hiring is the wedge. The platform is the destiny.

A production-ready AI interview engine today, built to become the conversational assessment infrastructure for every industry.