Run technical interviews you can’t technically run.
Interviewary is a side-panel copilot for non-technical recruiters. As the candidate speaks, it rates each answer, tells you why, and suggests the follow-up that probes real depth — then hands you a hiring report at the end.
Free · local-first · bring your own keys · ~$0.09 per interview
“…so I’d reach for a database index on the lookup column. It turns the full table scan into a B-tree seek, so reads stay fast even as the table grows…”
Names the right mechanism and the trade-off. Correctly connects the index to query performance at scale — not a memorized buzzword.
“What’s the downside of adding that index — what does it cost you on writes?”
Probes depth: a strong answer mentions slower inserts / index maintenance.
A recruiter can hear the answer. They just can’t tell if it’s any good.
Startups run technical screens with whoever’s available — a recruiter, an ops lead, the founder. They can read the questions and nod along. But they can’t tell a senior answer from a memorized one, or decide what to ask next. So the wrong people pass the screen and waste an engineer’s time in round two.
- ✕Every answer sounds equally plausible.
- ✕“Was that a good answer?” goes unanswered.
- ✕No idea what to ask next to test for depth.
- ✕False positives bounce back from engineering.
- ✓Each answer gets a rating + a one-line reason.
- ✓You see exactly why it was strong or weak.
- ✓Smart follow-ups appear in ~2 seconds.
- ✓A hiring report your engineers will trust.
From “hello” to a hiring decision, without leaving the call.
Upload & configure
Open the side panel in Google Meet, drop in the job description and résumé, pick seniority and interview type. You get 5–7 tailored opening questions — each with a note on what a strong answer sounds like.
Ask and watch
Run the interview normally. Interviewary transcribes both sides, splitting interviewer from candidate, so you never lose the thread of the conversation.
Read the answer, instantly
The moment the candidate stops talking, you get a rating — Weak, Adequate, Strong, Exceptional — with a one-line reason, plus 2–3 follow-up questions that probe what they just said.
Download the report
Click End interview and get a markdown report: summary, every Q&A with its evaluation, strengths, concerns, and a hire / no-hire recommendation — ready to share with the team.
Everything a non-technical interviewer needs — and nothing they don’t.
Real-time answer evaluation
A rating and a one-line reason appear seconds after each answer — written for someone who doesn’t know the tech.
Smart follow-up questions
Contextual probes based on what was just said, so you can tell real understanding from rehearsed buzzwords.
Tailored starting questions
Upload the JD and résumé to get opening questions matched to the role and the candidate’s seniority.
Live diarized transcript
The full conversation, transcribed in real time and split by speaker. Nothing gets lost.
Auto hiring report
A shareable markdown write-up with Q&A, evaluations, strengths, concerns, and a clear recommendation.
Local-first & bring-your-own-key
No accounts. Your transcripts, résumés, and keys stay in your browser — interview content never touches our servers. Usage analytics are opt-in.
Private by design. Fast by necessity. Yours to control.
Your audio, transcripts, résumés, and keys never leave your browser. Usage analytics are opt-in and anonymous.
Lives in the Chrome side panel using native tab capture — no second app, no window switching.
Anthropic, OpenAI, Google, and xAI. Pick your provider and models; you’re never locked in.
A strict latency budget keeps ratings and follow-ups fast enough to keep the conversation flowing.
You pay your own usage-based API costs. No subscription, no per-seat pricing, no markup.
Set up in under five minutes.
There’s no account to create, and your keys and interview data stay on your device. Download the extension, load it into Chrome, paste your own API keys once, and you’re interviewing.
Not on the Chrome Web Store yet — this is an early prototype, so it loads as an “unpacked” extension. The steps on the right walk you through it.
- 1
Download & unzip
30 sGrab the extension bundle and unzip it anywhere you’ll keep it — Chrome loads the extension from this folder, so don’t delete it afterward. - 2
Load it into Chrome
1 minOpen the extensions page, turn on Developer mode (top-right), click Load unpacked, and select the unzipped folder.Paste this into your address bar — Chrome won’t let pages link to it directly. Then pin Interviewary so it’s one click away.
- 3
Add your keys (one time)
2 minOpen Google Meet, open the side panel, hit the gear → Settings, then paste:- •Your Deepgram key (transcription — comes with $200 of free credit).
- •An LLM provider + key — Anthropic, OpenAI, Google, or xAI — then pick a fast model (live) and a report model (end of call).
Click Test keys to confirm they work, then Save. Keys are remembered on your machine — you only do this once.
- 4
Start interviewing
1 minBack on the setup screen, drop in the job description and candidate résumé, choose seniority (Junior / Mid / Senior) and type (Screening / Deep dive), and hit Start interview. Ratings and follow-ups start flowing as soon as the candidate talks.
Questions, answered.
Where does my data go?
Your interview content stays with you: audio, transcripts, and résumés are processed in your browser and sent only to your own Deepgram and LLM accounts, and your keys live in local storage. The only things we collect are the email you give us to download and — if you opt in — anonymous usage stats (never transcripts or answers). Full details are in our Privacy Policy.
What does it cost?
The extension is free. You pay your own usage-based API costs, which come to roughly $0.09 per interview for the LLM plus transcription — and Deepgram includes $200 of free credit when you sign up.
Which AI providers can I use?
Anthropic (Claude), OpenAI (GPT), Google (Gemini), and xAI (Grok). You choose a fast model for live ratings and follow-ups, and a higher-quality model for the end-of-call report.
Does it work on Zoom or Teams?
Not in this prototype — it’s built specifically for Google Meet, where it captures the call audio from the browser tab.
Why isn’t it on the Chrome Web Store?
It’s an early prototype, so you install it as an unpacked extension — download the bundle, then “Load unpacked” in Chrome. The setup section above walks through every step.
Do I need to be technical to use it?
No — that’s the whole point. Every rating, follow-up, and report is written for someone who doesn’t know the underlying technology. The one technical-ish step is pasting in API keys, which the guide covers.
Running into trouble — like the extension not picking up audio? Visit the Help & FAQ for setup and troubleshooting.