Defend Pre-Screen Risk Signals for Lever – Proof Help Center
Defend Pre-Screen Risk Signals for Lever is an automated feature that, when enabled, uses candidates' existing PII in Lever to run a risk assessment via Proof's Defend engine at a configured hiring stage, then writes a risk classification note and tag (low, medium, or high) back to their Lever record within minutes, allowing recruiters to filter and triage candidates based on these directional risk signals directly inside the ATS without requiring additional candidate input.
The short answer: Defend Pre-Screen automatically runs a risk check on candidates in Lever when they reach a configured hiring stage, then writes results — a risk classification and tag — directly back to their Lever record so your team can triage without leaving the ATS.
How It Works
When a candidate advances to a configured stage in Lever, Proof automatically pulls their available PII — name, email, phone, and address — and runs it through the same Defend risk engine your organization uses for transaction screening. No action is needed from the recruiter. Results are written back to the candidate's Lever record within minutes as a note and a tag (proof_defend_low, proof_defend_medium, or proof_defend_high), so your team can filter and act on risk signals right inside Lever.
The pre-screen uses only the PII already present in Lever — no selfie, no credential upload, and no action required from the candidate.
Pre-screen results are a directional risk signal, not a definitive determination. Use them alongside your normal hiring judgment — a medium or high tag does not automatically disqualify a candidate.
What Gets Written Back to Lever
After screening runs, Proof adds the following to the candidate's Lever record:
- A note — a summary of the risk signals evaluated, including any flags on phone, email, or identity data.
- A tag — one of proof_defend_low, proof_defend_medium, or proof_defend_high, based on your organization's configured Defend thresholds.
You can filter your Lever pipeline by these tags to quickly surface candidates that warrant a closer look.
Pro tip: The risk engine uses the same thresholds set for your organization's Defend configuration. If you want to adjust sensitivity, reach out to your CSM or Proof Support.
Requirements and Setup
This feature is off by default and must be enabled for your organization. To get started, you'll need:
- Defend enabled on your Proof account — this feature is included at no additional cost as part of your Defend subscription.
- A Lever ATS account with admin access to configure integration settings.
- A Lever API key — provided by your Lever admin and shared securely with Proof during setup.
Once enabled, your CSM or Proof's implementation team will configure which Lever stage triggers the screening and connect the integration using your credentials. You won't need to make changes in Proof's portal — everything surfaces in Lever.
This feature is available to Enterprise and Commercial customers using Lever. Contact your CSM to request enablement.
Summary Checklist
- Defend Pre-Screen runs automatically when a candidate reaches a configured Lever stage — no recruiter action needed.
- Results appear in Lever as a note and a proof_defend_low/medium/high tag.
- Candidates experience zero friction — no selfie or document upload required.
- The feature is off by default and included in your Defend subscription — contact your CSM to enable it.
Still unsure? Contact Proof Support for help.
Related
Proof Defend Overview
Proof Defend is an AI-powered fraud detection tool that analyzes transactions using identity and behavioral data—such as email, phone, and location—to provide real-time risk scores, detailed identity reports, deepfake video detection, and alerts for suspicious activity, with a free 30-day trial available directly through the Proof portal.
2026-07-27 Proof Release Notes DEPLOY-1324
The July 27, 2026 Proof release introduces a configurable rules engine for Defend fraud detection enabling dynamic global and customer-specific risk policies managed internally by Proof, an AI-generated plain-language risk summary in the Defend Risk tab and downloadable reports with user feedback options for Owners and Admins, and manual document template selection during uploads now available to Business customers upon request, enhancing fraud detection flexibility, transparency, and document handling efficiency.
Proof Q3 2026 Product Roadmap
The Proof Q3 2026 Product Roadmap highlights recent launches including Verified Business for trusted organizational profiles, Document Template Application for reusable document tag sets, Risk Decision Summary providing AI-generated risk assessments, and VC Issuance & Presentation API for cryptographically-signed identity verification, with upcoming integration into the Rippling platform to embed identity verification within HR workflows.
Introducing Defend | Proof
Proof, a pioneer in online notarization, has developed Defend, the first signature platform with built-in fraud detection designed to continuously verify customer identity and protect businesses from costly forgery, impersonation, and falsified records in critical transactions such as wiring money and transferring property titles, addressing an $81 billion identity theft problem in the US.
Defend – Sophisticated Fraud Intelligence Platform
Defend is a sophisticated fraud intelligence platform that leverages over a decade of aggregated data, AI-driven deepfake detection, explainable AI, and network signals to monitor and secure every customer transaction—ranging from signatures and notarizations to video-based identity verifications—by flagging suspicious activities for review and providing a comprehensive 360° view of customer identity to protect against diverse fraud types including forgery, impersonation, and coordinated attacks.
Introducing Defend's New Fraud Model
Defend is Proof's advanced enterprise-grade fraud prevention model designed to protect critical real-time transactions by leveraging over a decade of diverse, high-value data and multi-signal inputs—including ID verification, behavioral, and consortium data—offering superior precision and adaptability compared to legacy models to minimize fraud and false declines during high-stakes digital interactions.