No such thing as a good hire in the abstract.Read the manifesto
Superposition Logo
Company soul

Tune the role against real decisions.

Calibration lines a live role up against people you actually hired and passed on. The agent learns your bar from examples, not adjectives.

See how it works

How it works.

Start from real examples

Pull past hiring decisions from your ATS and calibrate the role against them, so the agent sees what a yes and a no look like at your company.

Past decisions, pulled in
Past decisions, pulled in
Your hires and passes come from the ATS so the agent sees real yeses and nos.
Rate a few examples
Rate a few examples
Approve or pass on examples and the agent works out what separated them.

Recalibrate without restarting

Needs shift mid-search. Re-tune a running role and the agent adjusts sourcing and FitCheck without starting over.

Re-tune a live role
Re-tune a live role
The role changed mid-search. Recalibrate and keep the pipeline you already have.
What moved
What moved
Weights shift and you can see exactly which ones.

Its own space

Calibration lives separately from onboarding, so you can return to it any time the role drifts.

Its own tab
Its own tab
Calibration sits apart from onboarding so you can return whenever the role drifts.
Decisions, dated
Decisions, dated
Every calibration decision is logged, so the bar is traceable rather than vibes.

You set the bar

The agent proposes, you decide. Calibration only records the decisions you make, and every one of them sharpens the next batch.

Setup guide in the docs

Discover more features

Voice intakeDeep web sourcingFitCheckBring your ownPersonalized outreachReplies handledReview queueGrace periodEscalationsAgent schedulingSignalInterview guidesScorecardsInsightsSlackMCP serverATS submissionCalendar and mailboxAPI and security

Get started

Find your perfect weirdo

The person who would fail anywhere else and belongs with you. Tell us the role and we will show you who the agent would go after first.

Start a search