✦ Crit DeckDemo deck Start How it works Sign in

★ INTERVIEWER COPILOT · TCG EDITION

Build a sharper
interview in minutes.

Paste a resume + JD. GPT-5.4 (high reasoning) deals you a 10-card pack — snapshot, expected answers, follow-up trees, a turn-meter timeline, and a guided level-up scoring wizard. No recording. No candidate-facing UI.

▶ START — PASTE RESUME + JD ✦ TRY A DEMO DECK

no login · runs on resume content only · ~3–5 min on high reasoning

MODULES PER PACK

CARDS DEALT PER MATCH

8 ✦

RUBRIC STATS

GPT-5.4

HIGH REASONING

EVIDENCE

RESUME · 14 HITS

✦

STR: 6 · RSK: 3
AMB: 4 · DIVE: 1

⚡PWR 60

QUESTION

SYSTEMS DEEP-DIVE

“take it 1× → 10× tomorrow — which three things?”

⚡⚡PWR 70

✦✦✦✦

WHAT YOU GET

four loops, one match

PACK

10-MODULE BRIEF

Snapshot, resume highlights, role-fit map, validate areas, question bank, expected answers, follow-up trees, rubric, red flags, timeline.

ARENA

BATTLE WORKSPACE

Three cards at a time. Mark asked, pin, log notes. Expected answers collapse so you don't anchor on them mid-conversation.

MAP

BIRD'S-EYE OVERVIEW

All cards in one view — filter by category, see asked / pinned, switch to a different question without losing your place.

LEVEL UP

GUIDED SCORING + REPORT

8 rubric dims · multiple-choice helpers · 1-5 score override · auto-drafted Markdown report. Copy, paste into Slack, done.

BUILDING DECK

fill 4 slots · open the pack

✦ need inspiration? try a demo deck →

Load the cards before opening the pack.

Paste a resume + JD (real ones — the model needs evidence). Lens is optional but improves expected-answer quality. We don't log resume text server-side.

SLOT 1SETUP● SETUP

CANDIDATE NAME

ROLE

TYPE

DURATION

LANGUAGE · questions, expected answers, report

SLOT 2CANDIDATE RESUME○ NEED ≥30 words

DROP A FILEPDF · DOCX · TXT · MD · ≤10MB · or just paste belowpick file

0 words · need 30 more

SLOT 3JOB DESCRIPTION○ NEED ≥20 words

DROP A FILEPDF · DOCX · TXT · MD · ≤10MB · or just paste belowpick file

0 words · need 20 more

SLOT 4INTERVIEWER LENS◐ OPTIONAL

tip drop 1–3 sentences. The model uses them to angle every question.

DECK READINESS

◐

Setup

name + role required

◐

Resume

0 words · need 30

◐

Job Description

0 words · need 20

◐

Interviewer Lens

optional

EXPECTED PACK

10 modules · 10–12 question cards · 8 rubric stats · follow-up trees · 1 level-up report

still need: candidate name · role title · resume (≥30 words) · job description (≥20 words)

privacy: resume + JD only used per-request; never logged server-side. Session lives in your browser's localStorage.

HOW IT WORKS

3 stages · zero login · all in your browser

OPEN THE PACK

Drop resume + JD. 10 cards reveal one by one in an intelligence pipeline — snapshot, highlights, questions, expected answers, follow-up trees.

PLAY YOUR HAND

During the interview, see 2-3 question cards at a time. Switch, pin, mark-asked. Expected outlines stay collapsed so you don't lead the witness.

LEVEL UP

End match → 8-stage scoring wizard. MCQ helpers nudge a 1-5 score per dim. Override. Then GPT-5.4 drafts the Markdown report from your notes.

SAMPLE MATCH

this is the Battle Arena you'll be in once the pack opens

OPPONENT ZONE

FE / AI hybridLV.54

MAYA CHEN

Sr. FE · AI Products

ATK 240DEF 180HP 800

Passive · Streaming First
System-design questions cost −1 ⚡.

ACTIVE LINE-UP · 3 cards selectable

★ FEATURED · Q.04⚡⚡PWR 70

SYSTEMS DEEP-DIVE

“You mentioned leading the migration of the editor to a streaming-first architecture. If you had to take it from 1× to 10× concurrent users tomorrow, which three things would you change first, and why?”

▾ EXPECTED⇡ FOLLOW-UPS · 3🎯 SYS · OWN

● ASKED · 📌 PINNEDQST/0407

NOW PLAYING

▾ EXPECTED ANSWER · Q.04

STRONG Names a specific bottleneck (CPU / socket / state sync).

⇡ FOLLOW-UP · 3 branches

if vague → which decision was hers?
if AI → how did she verify?

YOUR PARTY · RUBRIC STATS

Role fit4/5

Tech depth4/5

Ownership3/5

Problem—