Your hiring pipeline, run by AI you already pay for.
InterviewDashboard screens every resume against every open role, builds your interview rubric, and scores each answer with verbatim evidence — then files the finished feedback form for you. AI runs through your own Claude, ChatGPT, or Kimi subscription: no API keys, no per-call fees, no cloud.
- No API keys
- No per-call AI fees
- Local-first — your data stays yours
- Windows & macOS
Runs on the AI subscriptions you already have
Claude Pro / Max
ChatGPT Plus / Pro / Team
Kimi membership
Folders in. Feedback forms out.
Everything in between is automated.
InterviewDashboard turns your file system into the integration layer. No ATS migration, no uploads — drop files in the client's folder and the pipeline moves.
Drop a resume
A watched folder per client ingests every resume the moment it lands — PDF or Word, text extracted automatically.
Resumes/AI scans the fit
Each resume is scored 0–100 against every open role for that client — verdict, summary, pros and cons, in seconds.
strong fit · 84Interview & drop the transcript
Interview from your own script. The transcript auto-matches to the right candidate and round by filename.
T2_ Dana Kim - Staff Engineer.pdfReview, finalize, file
AI scores every answer against your rubric with verbatim evidence. You edit, finalize, and export a formatted feedback form — auto-filed.
Interview Feedback Forms/Built for teams that interview
at enterprise volume.
Every resume, ranked against every open role. Automatically.
The moment a resume hits a client's watched folder, InterviewDashboard extracts the text and runs a fitment scan across all of that client's open roles — a 0–100 fit score, a verdict, and the reasoning, before you've even opened the file.
- ✓Strong fit / good fit / weak fit / not a fit verdicts calibrated to your rubric language.
- ✓Pros & cons per role, plus a resume summary with experience, key skills, and gaps worth probing.
- ✓One click to candidate — promote a scan straight into the pipeline, no re-typing.
8 yrs distributed systems; led payments platform rebuild at 40M req/day scale.
- ✓Deep Go / Kubernetes match
- ✗No fintech domain time
- ✓Team-lead experience
- ✗Contract-heavy last 2 yrs
Your interview script becomes a scoring rubric. Versioned.
Upload the script your interviewers already use. AI extracts weighted evaluation criteria and writes good / better / best reference answers for every question — so every interviewer scores against the same bar, and every score means the same thing.
- ✓4–7 weighted criteria, weights summing to 100 — tuned to the role.
- ✓Tiered reference answers: what a 3, a 4, and a 5 look like, in your firm's own language.
- ✓Fully editable & versioned — refine the rubric, keep the history.
Working token-bucket design, one datastore, basic trade-offs.
Distributed design, Redis vs in-memory, failure modes called out.
Adds fairness, burst policy, observability, and rollout plan unprompted.
Every answer scored. Every score backed by the transcript.
A two-pass review maps each scripted question to the candidate's actual answer, then scores Quality and Accuracy 1–5 against your tiers — with a verbatim quote as evidence. You get a weighted overall score and a hire recommendation you can defend to the client.
- ✓Evidence-anchored scoring — no score without the quote that earned it.
- ✓Strengths, gaps & concerns, plus what a best answer would have covered.
- ✓Human-in-the-loop: edit any score before you finalize and export the .docx feedback form.
Consistently strong, structured answers with real production depth. System design answers reached the best tier unprompted.
- ✓Rate-limiter design hit best tier
- ✗Shallow on testing strategy
- ✓Owned a migration end-to-end
- ✗No multi-region exposure
Organized the way recruiting firms actually work.
Clients own roles; roles own candidates and interviews. Each client gets a clean folder tree on disk, and InterviewDashboard keeps it fed — resumes in, transcripts matched, feedback forms filed. Your files stay readable, portable, and yours.
- ✓Multi-client by design — run every account from one dashboard, cleanly separated.
- ✓Role pipeline: draft → open → on hold → closed; candidates flow applied → hired.
- ✓Finished feedback forms auto-filed to the client's folder, named and dated.
Serious about cost. Serious about
candidate data.
InterviewDashboard is a desktop app, not another SaaS meter. The architecture choices that make it cheap to run are the same ones that make it safe to deploy.
Subscription-billed AI
AI work is driven through the Claude Code, Codex, or Kimi CLI already signed in on the machine. Usage bills to the subscription you already pay — InterviewDashboard never touches an API key.
per-call AI fees: $0.00, forever
Local-first privacy
Candidates, transcripts, scores, and reviews live in a local database and plain folders on your machine. Resume text goes only to the AI vendor you signed in with — never to us.
cloud accounts required: none
Per-task model routing
Route each of the seven AI task types to the provider, model, and effort level you choose — cheap models for matching, flagship models for scoring. Every run is logged with tokens and duration.
audit trail: every prompt run, on record
Human-in-the-loop
AI proposes; the interviewer disposes. Every score, summary, and recommendation is editable before finalize — and the exported feedback form credits the reviewer who signed it off.
final say: always the interviewer
One license. One price.
No meter running.
You already pay for AI. InterviewDashboard makes that the only AI bill you'll ever see.
One-time purchase. Yours to run, seat by seat.
- ✓Unlimited clients, roles, candidates & interviews
- ✓Resume inbox with automatic fit scans across every open role
- ✓AI rubric builder with good / better / best tiers
- ✓Evidence-scored interview reviews & .docx feedback forms
- ✓Per-task model routing + full prompt-run audit log
- ✓Windows & macOS desktop apps
Requires an active Claude Pro/Max, ChatGPT Plus/Pro/Team, or Kimi membership on the machine. That's the only ongoing cost.
The math your CFO will like.
No usage meter
Screen 10 candidates or 10,000 — AI runs on your flat-rate subscription, so heavy interview months cost the same as light ones.
No per-seat platform tax
$250 once per license. No monthly SaaS fee scaling with headcount, no candidate-volume tiers, no overage surprises.
Volume licensing for firms
Outfitting a whole recruiting team? Talk to us about firm-wide licenses.
Questions, answered.
No. InterviewDashboard drives the Claude Code, Codex, or Kimi CLI already installed and signed in on your machine. There are no API keys to manage, no usage dashboard to watch, and no per-call charges — AI usage simply bills to your existing subscription.
Claude Pro/Max (via Claude Code), ChatGPT Plus/Pro/Team (via Codex CLI), and Kimi membership (via Kimi Code). You can use one or all three, and route different task types — resume scans, rubric generation, interview scoring — to different providers and models.
On your machine. Candidates, roles, transcripts, scores, and reviews are stored in a local database and plain folders you control. There are no InterviewDashboard cloud accounts. When AI runs, document text goes only to the AI vendor you're signed in with — never to us.
It's built to be the layer recruiters actually live in. Folder watching, candidate pipelines, and auto-filed feedback forms mean many firms run it instead of an ATS for interview operations — and because everything is plain files and folders, exporting into any system you keep is trivial.
InterviewDashboard is a native desktop app for Windows and macOS. It's fast, works offline between AI runs, and feels like part of the OS — because it is.
One license covers one installation of InterviewDashboard with unlimited clients, roles, candidates, and interviews. It's a one-time $250 purchase — no subscription for the software itself. Your AI subscription (Claude, ChatGPT, or Kimi) is the only ongoing cost.
Your next hire's feedback form
could file itself.
ONE-TIME LICENSE · NO API KEYS · WINDOWS & MACOS