picked.
Log inPost a role
About · the story so far
Picked.ai · since 2023
We started in 2023.
Picked.ai began in 2023 with a simple bet: hiring should be a science, and the slow, manual parts of it should be handled by an AI that does not get tired, rushed, or biased. We have been building ever since, and we have shipped over 3 million assessments along the way. Picked.ai screens, assesses, and interviews every applicant, then returns a ranked shortlist in 48 hours.
2023
Picked.ai founded
3M+
Assessments shipped
43k
Items in our library
48 hrs
Role to shortlist
Timeline
Built since 2023. Now open to everyone.
2023
Picked.ai founded
A validated psychometric item bank, a frontier model, and one bet: an AI-native pipeline that screens, assesses, and interviews every applicant, then returns a ranked shortlist. We started building.
01 / 04
2024
The frontier-model moment
Anthropic's Claude 3 family made the conversational screen feasible. Candidates could speak instead of type, and the score still held up against the validity data.
02 / 04
2025
Internal pilot
48 friendly customers ran roles end to end through a closed beta. Two hires per role on average, sourced through us instead of around us. Median time-to-shortlist dropped from 32 days to 31 hours.
03 / 04
2026
Picked.ai opens to the public
Self-serve. 99c per vetted candidate. First fifty free. The full pipeline, with AI doing the parts no manager wanted to do.
04 / 04
The moat
Anyone can call an LLM. Nobody else has the item bank.
↓
A validated item bank, calibrated against real hire outcomes
Every question we ask a candidate is calibrated against whether the candidate ended up performing in the role. The LLM is the interface; the bank is the substance.
↓
Psychometric scoring, not vibes
Each response is graded against a rubric that has been peer-reviewed and published. Not "the model thought they sounded smart". A specific, defensible scoring matrix.
↓
Cheat-resistance built in, not bolted on
Voice, adaptive item ordering, and response timing all roll into the score, so a leaked typed answer does not break the assessment.
↓
Built to last, not to flip
We are bootstrapped and not racing to spend a Series A. We can make the long bet on the candidate side of the marketplace because we don't need this quarter's revenue.
Picked.ai in one line
AI-native hiring
Since 2023 · London
Describe a role. Picked syndicates it, screens every applicant, runs an assessment, conducts a first-round interview, and returns three finalists in 48 hours. 99c per vetted candidate, no seats, no minimums.
Built on a validated psychometric item bank. Over 3 million assessments and counting.
Team
The team. Building Picked.ai.
See open roles →
Team member
Co-founder · CEO
Profile coming soon.
Team member
Co-founder
Profile coming soon.
Team member
Head of fairness
Profile coming soon.
Team member
Head of engineering
Profile coming soon.
Team member
Head of growth
Profile coming soon.
Team member
Head of design
Profile coming soon.
What we believe
Hiring is a science.
We have the data.
The Picked principle
The market has overpromised on AI hiring for ten years. We are not interested in being part of that. We want a product that, in five years, hiring managers say works, and that legal teams say didn't surprise them. The first part is the product; the second part is the company behind it.
Backed by
Bootstrapped
since 2023
Founder-owned
100%
Not raising
this year
We have been profitable and bootstrapped from the start. We're not racing to deploy a Series A. The product gets to take its time.
The bet
Three years of work.
Forty-eight hours of yours.
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.
About · Picked.ai