AI and credits
AI that assists. People decide.
Every AI feature in Panelynx helps you read a candidate; none of them makes the call for you. Each action runs on your workspace credits, and AI hands you a recommendation and its reasoning to weigh.
AI gives a second opinion. People make the call.
David Okafor
Senior Backend Engineer · Match
Core backend stack matches; owned a large migration end to end.
Lighter on the role domain. Worth a closer look in the interview.
A read on the fit, with its reasoning. You decide.
Second opinion
A read on the fit, after a panelist locks.
Ask AI for a second opinion on a candidate and it returns a couple of plain notes with its reasoning: what looks strong, what is worth a closer look. It runs after the human read, never instead of it, and you decide what to do with it. It is a signal, not a verdict.
- Two plain notes plus the reasoning behind them
- Requested after the read, never an auto-decision
- Runs on your workspace credits, never automatic
David Okafor
Senior Backend Engineer · Match
Core backend stack matches; owned a large migration end to end.
Lighter on the role domain. Worth a closer look in the interview.
A read on the fit, with its reasoning. You decide.
Resume parsing
Drop a CV; the fields fill a draft you correct.
Upload a resume and AI reads it into the candidate record: name, title, location, experience, education, and skills. It fills a draft, not a final answer, so you skim it and fix anything it got wrong.
- Name, title, experience, education, and skills
- A draft you correct, never a locked answer
- Runs on your workspace credits
Parse resume
Fills a draft you correct
Parsed from the PDF
A draft you correct. You stay in control.
5 cr from a PDF · 1 from textQuestion generation
AI drafts questions. You keep or edit.
Ask AI to draft interview questions from the role and the sections you have already set, then decide on each one. It hands you a starting point; you keep what fits, edit the rest, and discard the rest. Nothing lands in your plan until you keep it.
- Drafts from the role and your existing sections
- Keep, edit, or discard every suggestion
- Runs on your workspace credits
Draft questions
From the role and your sections
Drafted by AI
Design a write-heavy event ingestion pipeline
Walk me through how you would shard a growing Postgres table
Describe a production incident you owned end to end
You keep what fits and edit the rest.
What's inside
Assistive AI. People decide.
AI hands you a recommendation and its reasoning. No predictions, no scoring people for you, and a person makes every call.
Metered credits
AI actions run on credits, so you pay for what you use.
Pooled monthly credits
Each plan includes a monthly pool of credits, shared by the team.
Top-up packs
Add credits when you need them; a pack beats per-credit overage.
Hard cap on overage
Overage is opt-in, priced per credit, and capped so spend can never run away.
Transcript read
Upload an interview transcript and AI reads it against your scorecard.
Assistive, never deciding
AI hands you a recommendation and its reasoning. A person makes the call.
Good to know
Questions, answered plainly.
No. Every AI feature in Panelynx is assistive: it hands you a recommendation and the reasoning behind it, and a person makes the call. The second opinion runs after a panelist has done their own read, parsing fills a draft you correct, and generated questions only land in a plan when you keep them. AI never finalises anything for you.
Start now
Run your next interview on the same standard for every candidate.
A complete applicant tracking system, postings, pipeline, scheduling, and offers, with structured interviewing built into the core. I'm early, by design, and I'm building it with a short list of teams who care about getting this right. It's a narrow door on purpose.
14 days free. No card.