picked.
Log inPost a role
Product · Shortlist
Three recommended finalists. The whole list behind them. Every score explained.
48 hours after you post the role. Open the email. Make the call. Or drill in and disagree; the model will re-rank with reasoning.
Post a role →
What lands in your inbox at 48 hours
SUBJECT: Three finalists are ready · Senior Backend Engineer · London
SENT: Tuesday, 06:42
app.picked.ai / roles / senior-backend-engineer / shortlist
Today · shortlist ready
Senior Backend Engineer
48 VETTED · 12 SHORTLISTED · 3 IN FINAL ROUND · $0.00 OF $50 CAP
Live23h 41m
MR
Marta Reinholt
Berlin · ex-Lyft
91/100
Top systems-design score across the cohort. Eventual consistency, no hand-waving. Direct, picks the right tradeoff.
TOP PICK · OPEN→
DT
Dario Tedeschi
Milan · ex-Spotify
88/100
Strong reliability instincts. Has shipped this exact system before. Lower bandwidth, asks one careful question.
OPEN PROFILE→
AY
Aiyana Yazzie
Remote · MST
86/100
Self-taught path, distinct from the other two. Pragmatic about tooling. Lower formal-CS depth but bright in the moments that count.
OPEN PROFILE→
9 more in the ranked shortlist · scores 78 - 85
The dashboard is the source of truth. The email is a nudge.
Drill in
Six things you see on every finalist, in one click.
01
Score breakdown.
Per-trait, with evidence. Not a single black-box number; a stack of numbers, each tied to a thing the candidate did or said.
02
Screen transcript.
The full conversation, scrollable, with the bits we found most informative highlighted.
03
Assessment results.
Per-module scores, percentile against role norms, and the raw responses on open-ended items.
04
Interview transcript.
Full audio and text, with the model's notes on each behavioural anchor and scenario.
05
Salary and notice.
Comp expectation, notice period, right-to-work, location, working pattern. All asked in the screen, all confirmed in the interview.
06
Suggested human-interview questions.
Three to five questions tailored to what we saw, sized for a 45-minute final round.
Side by side
Compare up to four candidates at once.

Pick two, three, or four candidates and Picked lays them out side by side. Strengths, risks, evidence, score breakdown, salary expectation, notice period, all in one view. Sortable by any trait. Exportable as a PDF for a hiring conversation with your team.

Compare · 4 of 12 shortlistedExport PDF ↓
AO
Amara Okafor
94/100 · top pick
Systems
96
Pragmatism
92
Ownership
95
Communication
91
Salary$135k
Notice4 weeks
TL
Theo Lindqvist
91/100
Systems
97
Pragmatism
88
Ownership
90
Communication
86
Salary$128k
Notice8 weeks
PA
Priya Anand
89/100
Systems
88
Pragmatism
90
Ownership
87
Communication
94
Salary$140k
Notice2 weeks
MW
Marcus Webb
85/100
Systems
84
Pragmatism
89
Ownership
82
Communication
88
Salary$120k
Notice4 weeks
Sample · illustrative data
Re-rank · Senior Backend Engineer
02Theo Lindqvist91/100
👎 You disagreed
Model · re-ranked with reasoningSystems-design weighted down for this role; stakeholder communication weighted up. Two positions changed. Theo's deep-infra signal is still strong; it now counts for less here.
01Amara Okafor
94
02Priya Anand
↑ up 190
03Theo Lindqvist
↓ down 188
Override · this role onlyApply to all future roles
Sample · illustrative data
When you disagree with the model
Thumbs down. The list re-ranks with reasoning.

Picked makes a recommendation; you do not have to take it. Thumbs down on any of the top three and the model re-ranks the list, explaining what it weighted differently and which signals it now treats as less reliable. You can over-ride permanently or just for this role.

  • ●Per-role calibration: your taste, not the platform's.
  • ●Per-trait override: tell us cognitive ability matters less for this role; we listen.
  • ●Per-criteria veto: hard-no on a specific signal; we promote a different finalist immediately.
What you do from the dashboard
Four buttons. The rest of the process flows from them.
Shortlist.
Pull a candidate into your human final round. Picked sets up the scheduling in V2; today you copy the candidate into your calendar.
Reject (respectfully).
Picked sends the candidate a personalised post-rejection note with feedback. You never write a polite no email again.
Message.
Ask the candidate a question. Their answer feeds back into the file.
Flag for human review.
Internal escalation if something feels off. A Picked reviewer audits the file within one business day.
48 hours
Median from posted to shortlist, across the beta.
3 of N
We recommend 3 finalists; you see all N.
100%
Every candidate gets a respectful, personalised decision. Zero ghosting.
You
make the hire. The model recommends; the decision stays with you.
Other product surfaces
PRODUCT · DECISION ENGINE
Every score explainable.
Glass box. Bias monitoring. Continuous validation.
Read →
PRODUCT · ASSESSMENT
The Neuroworx item bank, ported to AI.
Cognitive, behavioural, situational, role-specific.
Read →
PRODUCT · CANDIDATE EXPERIENCE
The other side of the marketplace.
Real conversation. Real decision. Profile they own.
Read →
Want to see your shortlist?
Post the role. Wait 48 hours.
Post a role →Pricing
picked.
The AI-native replacement for inbound hiring. For the hiring manager, not HR.
© 2026 Picked · Made in London · Three finalists in forty-eight hours.
Shortlist · Picked.ai