VP Talent
Needs work-sample evidence legal and enablement can audit, without AI ranking candidates.
Revenue Command Center · Cluster
Hiring with evidence
For VP Talent, hiring managers, and CRO / GTM leads who need comparable work-sample evidence: seat-mapped requisitions, band thresholds, consent, and an auditable hire / advance / hold / decline record. Not AI auto-hire.

Hiring fails when AI ranks candidates or when recruit scores quietly pollute employee averages. Revenue Command Center records a recruit assessment contract, scores exceeds / meets / below / insufficient evidence bands, lets Klara explain evidence only, and requires a human decision with rationale.
Buyers
VP Talent
Needs work-sample evidence legal and enablement can audit, without AI ranking candidates.
Hiring manager
Needs clear exceeds / meets / below / insufficient bands and a place to record hire, advance, hold, or decline with rationale.
CRO / GTM lead
Needs seat-mapped hiring (SDR, AE, CSM) that does not pollute active-team capability averages.
Seat map
HIRESDR, HIREAE, and HIRECS all resolve to public room SDEALQ with a forced track and forcedPurpose recruit. Seat map stays intact: no fake CSM acronym, no Team IQ framework clones as hiring gates.
HIRESDR
Forced track sdr. Assesses SPICED handoff under recruit purpose. Demo seniority band experienced.
HIREAE
Forced track ae. Assesses MEDDPICC under recruit purpose. Demo seniority band enterprise.
HIRECS
Forced track cs. Expand/renew assessment under recruit purpose. Demo seniority band strategic_csm.
Contract
RecruitingAssessmentContract version recruit_v1 records roleTrack, hiringSeniority, jobBriefSnapshot (title, mustHaves, icp, success90d), scenarioPackId and scenarioPackRevision, rubricVersion, timeBoxMinutes, requiredEvidence, mustPassGates, comparisonCohortId, accommodations, and humanReviewPolicy. When a candidate starts, the system timestamps contractLockedAt. That is a recorded start lock, not a claim of immutable server-side prevention of later edits, and timeBoxMinutes is not a runtime-enforced timer.
SDR time boxes
Foundation 30 min, experienced 40 min, senior_lead 50 min.
AE time boxes
Commercial 40 min, mid_market 55 min, enterprise 70 min.
CS time boxes
CSM 30 min, senior_csm 45 min, strategic_csm 60 min.
Required evidence (all bands)
Authored decision rationale, scenario evidence citations, and live interaction record. Senior bands add track-specific artifacts.

Bands
Literal bands are exceeds, meets, below, and insufficient_evidence. Workspace ordering is exceeds, then meets, then below, then insufficient, then name. Overall competency uses worst scored band wins when any competency meets or exceeds; if every scored competency is insufficient and none meets or exceeds, overall stays insufficient.
Exceeds
Score at least 85.
Meets
Score from 70 inclusive to under 85.
Below
Score under 70.
Insufficient evidence
Score is null or NaN, or no competency items exist to score.
Klara
In recruit mode the candidate experience suppresses coaching UI and desk tips. In the RCC workspace Klara can answer evidence questions. The server authenticates the manager, caps the prompt at 1,200 characters, records an audit entry, and refuses without calling the model for four categories.
hire_recommendation
Refuses: I cannot name who to hire.
rank_people
Refuses: I cannot rank candidates.
compare_people
Refuses: I cannot say who is better for hiring.
thin_context
Refuses when context is too thin; short prompts under eight trimmed characters get a clearer-evidence ask.
Allowed answers must stay on provided evidence, cite competency, moment, band, and contract revision, disclose thin evidence, and never hire, rank, compare, or name someone better.
Human decision
The decision endpoint accepts only hire, advance, hold, or decline. It requires manager ownership of the team and a rationale of at least 40 characters, then appends an auditable decision record with evidence references and optional contract fields.
Hire
Human chooses hire with written rationale.
Advance
Human advances the candidate to the next stage.
Hold
Human holds pending more evidence or process.
Decline
Human declines with written rationale.
Firewall
shouldApplyTrainingProgression for recruit returns false. When a session is recruit mode, EnablementService does not write capability signals and does not call skill-profile-write. Recruiting evidence must never train employee capability profiles or cohort aggregates. Managers may explicitly Start training from assessment to convert a completed candidate into a new learning baseline. That is a separate opt-in action, not automatic aggregation.
No silent profile write
Recruit sessions skip capability signals and skill-profile-write.
Explicit conversion only
Start training from assessment is a manager action, not a background merge.
Candidate
Consent version is recruit_consent_v1. A candidate cannot start unless metadata.candidate.consent.acceptedAt exists. Acceptance stores version, noticeVersion, acceptedAt, lifecycle consented, and retentionUntil. Default retentionUntil is 180 UTC days from the supplied date or now. Intake has an explicit forbidden list: race, ethnicity, gender, age, religion, disability, nationality, sexualOrientation. The candidate report allow-list includes candidateName, roleTitle, requisitionTitle, optional completedAt, competencies with label / band / evidenceRefs, humanReviewPolicy, and optional feedbackNotes. Ranking, peer identity, hiring decision, and peer-relative overall band fields are stripped.
Consent gate
No start without accepted consent at recruit_consent_v1.
180-day retention window
retentionUntil defaults to 180 days. This is a recorded policy window, not an automated purge worker.
Allow-listed report
Completed candidates can receive a tokenized results link that cannot leak ranking or hire decisions.
Boundaries
Claim-safe boundaries for legal, talent, and enablement reviewers.
Intake forbids demographic fields (race, ethnicity, gender, age, religion, disability, nationality, sexual orientation). Scoring uses work-sample bands on a recorded contract. Humans decide with a written rationale. We do not claim a bias-reduction percentage.
Bands still compute per candidate. Insufficient evidence stays visible. Klara still refuses ranking and comparison. Small pools do not unlock AI shortlists.
Compare on the same requisition contract: pack revision, rubric version, seniority, and must-pass gates. Bands are exceeds, meets, below, or insufficient. Humans still own hire / advance / hold / decline.
No. Team IQ may support. It is never a lone hiring gate. Work-sample bands and human decisions sit in the recruiting workspace.
Recruit purpose suppresses coaching UI for candidates and never writes employee capability profiles. Coaching and mastery queues apply to active team members after an explicit conversion, not during the hire assessment.
Evidence
Read the hiring seat page, then book discovery when you want a facilitated walkthrough of the recruiting workspace.
Open the manager evidence gallery, then book a discovery call when you want a facilitated walkthrough of Revenue Command Center.