Let’s talk

Products & Labs · Hiring intelligence

Insignia Recruit

Hiring intelligence, scored from public signal

v.1.8 · retiredRetired · shipped 2024–2026

A hiring engine that scored candidates against engineering competencies using public technical artifacts — commits, papers, talks — rather than CV keyword matching. Skill-from-signal beats skill-from-claim.

Read the buildView all products
Insignia Recruit — Hiring intelligence, scored from public signalInsignia Recruit · v.1.8 · retired

Overview / Hiring intelligence

Why we built it

Senior-engineering hiring was being conducted on CV scans.

Public signal — open-source commits, conference talks, technical writing — actually evidences competence, and it was abundant and largely ignored. Recruit inverted the priority: public artifacts became the primary evidence and the CV became supporting material.

  • Skill graph
  • Calibrated scoring
  • Bias-aware shortlist
  • ATS integration

For: Was built for talent teams at engineering-led organizations

Ask about the build

The build

What it shipped.

The studio’s first published product: scored candidates against engineering competencies from public technical signal. Lessons absorbed into Auspex.

01 / 04

Skill graph

Derived from public technical artifacts.

02 / 04

Calibrated scoring

Across roles and seniorities.

03 / 04

Bias-aware shortlisting

With audit trails on every decision.

04 / 04

Native ATS integration

With Greenhouse, Lever and Workday.

Lessons

What we learned.

And what we would do differently.

  1. 01

    Calibration is a discipline, not a feature

    Once calibration was treated as a first-class production concern, every downstream decision improved.

  2. 02

    Audit trails on every shortlist

    Recommendation-grade decisions that don’t carry their reasoning forward become noise.

  3. 03

    Bias-aware scoring is permanent

    A standing architectural constraint that gated every model change — not a launch task.

  4. 04

    Export-first, not adapter-first

    Three bespoke ATS integrations were maintenance-heavy; generated artifacts would have aged better.

47
Engagements across 12 organizations
89%
Accuracy against the eventual hire
12d
Median time-to-shortlist
0
Audit failures
Retired · 2024–2026

Why it’s retired.

Studio focus consolidated to flagship products. Recruit was retired honorably: existing accounts were migrated to alternative providers, and data was returned to operators on request.

See AuspexAll products

Where the lessons live now

  1. 01

    Calibration discipline

    Every Auspex recommendation carries calibration metadata. The pattern originated here.

  2. 02

    Audit-trail-first design

    The audit ledgers in Verum and Auspex trace their lineage to Recruit’s shortlist audit trail.

  3. 03

    Bias-aware scoring as constraint

    Now a permanent architectural posture across studio products.

Discover more

More from the Insignia portfolio.

Vertical SaaSv.1.5 · liveLynkedUp ProNext-gen roofing CRMA vertical operating system for roofing contractors: CRM, ERP, AI-assisted damage assessment, property scouting and pipeline-grade lead generation.For: Roofing contractors · regional and nationalExplore the dossier Clinical platformv.1.x · liveVerumClinical inference platform · payer + providerTwo production-grade clinical-decision surfaces on one deterministic-gate engine: Decide for payer prior auth, Insight for provider multimodal decision support.For: Payer utilization-management teams · provider clinical-decision groupsExplore the dossier Managed content imprintv.1.x · liveSigilManaged content imprint · direct or agencyA studio-run content operation on a per-brand retainer. Staff produce the posts in your voice; you review and approve — every social format on the calendar.For: Brand owners · agencies on retainerExplore the dossier

Build with the studio

Want the patterns behind it in your own system?

Read the buildAll products