Skip to content

Capability · the moat

Shortlists you can defend, line by line

A score with no reasoning is a liability. Trasor's matching engine ranks people against the agreed spec and attaches the evidence behind every result.

ExplainableBidirectionalHuman-in-the-loop

What makes it different

Matching that explains itself

Bidirectional

Both sides ranked

Candidates are scored against the role, and roles against the candidate's goals — so fit is mutual, not one-directional.

Evidence-linked

Reasons attached

Each match cites the evidence behind it — skills, work history, assessments — drawn from the candidate's Professional Record.

Auditable

Defensible by design

Every ranking is traceable and reviewable, with human override always in the loop. The moat incumbents can't copy with UI polish.

Under the hood

An architecture, not a single model

Matching is designed as selectable engines rather than one opaque score. The ranking quality is the core story — so it's instrumented, evaluated, and improved with real outcome labels.

  • Ranks against the structured request, not keyword soup.
  • Match outcomes are captured and labeled, so the engine keeps learning.
  • Surfaces near-matches and the specific gaps that hold a candidate back.
  • Every decision keeps a human reviewer in control.

match.result

Distributed systemsStrong

Strong · 3 evidence items

GoStrong

Strong · verified assessment

Event-drivenPartial

Partial · 1 project

KubernetesGap

Gap · no evidence yet

Rank on evidence, not vibes

See how an explainable shortlist reads when every result comes with its reasons.