Pre-hire assessments that survive a defensibility review

Truffle's assessments are positioned as decision-support, not as automated employment selection tools. Personality uses IPIP, a peer-reviewed public-domain Big Five instrument with documented reliability data. Situational Judgment and Environment Fit are diagnostic instruments without correct answers, no auto-rejection thresholds, and transparent scoring. Per-position job-relatedness rationale is captured at setup and archived alongside candidate responses. The architecture is designed for the regulatory direction, but compliance is jurisdiction-specific and depends on how you operationalize the tool. Confirm your specific posture with employment counsel for the states and countries where you hire.

Truffle uses IPIP, the International Personality Item Pool — a public-domain Big Five instrument developed by Lewis R. Goldberg and used in decades of peer-reviewed research. The Standard form (50 items, 10 per trait) typically reports strong internal-consistency reliability across the five traits. The Express form (20 items, 4 per trait) reports moderate reliability suitable for high-volume screening with appropriate caveats. The IPIP item pool and its psychometric history are publicly documented and citable.

No, and they're not positioned as such. Situational Judgment and Environment Fit are diagnostic instruments designed for interview conversation prep, not for scored gating. They have no correct answers, no auto-rejection thresholds, no validity claims that would push them into the formal psychometric selection tool category. Use them as alignment information for the next interview round, not as pass/fail decisions on their own. The diagnostic framing is consistent across product copy, candidate notices, and configuration UI.

No. There is no threshold-based rejection setting anywhere in the assessment flow. Every result lands as diagnostic data inside the candidate profile alongside resume parse output and (if used) video interview scoring. A human reviewer advances, holds, or rejects every candidate. That architecture eliminates the adverse-impact pattern that auto-rejection pre-employment platforms produce structurally.

Per-position aggregate exports support standard adverse-impact analysis across assessment outcomes, designed to align with the four-fifths rule reporting format that internal HR and outside counsel typically request. The export includes assessment results, role configuration, and decision outcomes per candidate. Combined with the no-auto-rejection architecture, the monitoring posture is built around making adverse-impact patterns visible to a human reviewer and an audit team rather than letting them accumulate inside an automated decision pipeline.

Configurable per-position notice copy, per-assessment consent capture, disclosure of what the tool measures, how the data is used, who reviews results, and how long data is retained. The flows are aligned with Illinois AIVIA notice and consent requirements, Maryland's facial-analysis consent law, NYC LL144 candidate-notice norms, and the EU AI Act's high-risk hiring classification. ADA accommodation requests and decline-and-continue paths are first-class so candidates aren't penalized for requesting accommodations or opting out of an assessment.

Per candidate: the response data per assessment, the alignment scores with reasoning, the role configuration that defined the scoring rubric, the job-relatedness rationale captured at intake, the consent record with timestamps, and (if used) the resume parse output and video interview transcript with AI Match reasoning. Exportable for adverse-impact review, EEOC inquiry, internal audit, candidate-record-request response, or plaintiff's counsel discovery. The package is assembled by default, not built on demand.

ADA accommodation requests are a first-class candidate flow. A candidate can request accommodation before starting an assessment, decline a specific assessment without losing their application, or use a decline-and-continue path that routes them to a manual review track. Accommodation requests trigger a manual review queue for your team. The flow is configurable per position and the language is editable so the disclosure matches your standard application copy.

Not effectively, because there are no universal correct answers. AI tools rely on knowing the right answer in advance. Truffle's assessments score against your specific role configuration, which an AI assistant has no access to. Both ends of every Personality trait carry legitimate strengths. Every Situational Judgment scenario has four genuinely defensible options. Both sides of every Environment Fit pair are equally valid working conditions. The gameability problem that breaks knowledge-based pre-hire tests doesn't apply here.

$149 per month on the Self-Serve plan, or $99 per month with annual billing. All three assessments included on every plan. Audit trail, consent and notice flows, ADA accommodation handling, and per-candidate archive are built into the standard product, not paid add-ons. 7-day free trial, no credit card required.