BrewShield

BrewShield is SkillBrew.AI's built-in assessment integrity and proctoring module – the technology layer that monitors candidate behavior during AI-administered assessments and interviews to ensure that evaluation results reflect the candidate's genuine capability rather than external assistance, answer sharing, or impersonation.

BrewShield is designed around a principle that most proctoring products ignore: assessment integrity and candidate experience are not opposing goals. Organizations that deploy invasive video proctoring – continuous webcam monitoring, eye-tracking, keystroke logging with aggressive flagging – produce lower completion rates, negative employer brand impressions, and candidate dropout that disproportionately affects candidates who are less comfortable with surveillance-intensive technology. BrewShield delivers meaningful integrity coverage without the candidate experience cost of maximum-intrusion proctoring.

Browser and Tab Activity

BrewShield detects focus events – when the candidate navigates away from the assessment window to another browser tab, application, or screen. Each focus event is logged with a timestamp. A pattern of frequent, brief focus events during answer-intensive questions is a strong indicator of answer-searching behavior.

Session Consistency

BrewShield tracks session metadata – device fingerprint, IP address, assessment timing patterns – that can identify anomalies suggesting impersonation (a different device connecting mid-assessment) or geographic inconsistencies that warrant human review.

Behavioral Timing Analysis

Unusually fast response times for complex questions, or suspiciously uniform response timing across candidates who might be sharing answers in real time, are flagged for review. Legitimate performance has natural timing variation; coached or copied responses often don't.

Copy-Paste Detection

For written response assessments, BrewShield logs copy-paste events – large blocks of text pasted rather than typed – which may indicate AI-generated or externally sourced responses.

AI-Generated Response Detection

For written and coding assessments, BrewShield's analysis flags response patterns consistent with AI generation – stylistic uniformity, unusual vocabulary consistency, structural patterns characteristic of large language model output – as integrity signals for human review.

BrewShield does not require:

• Continuous webcam video recording
• Room scanning before assessment begins
• ID document verification at session start
• Keystroke-level logging and playback

These measures are available as enhanced options for high-stakes assessments where integrity requirements justify the additional candidate experience cost – but they are not the default, because the default BrewShield configuration provides meaningful integrity coverage for the majority of hiring assessment contexts without requiring candidates to submit to surveillance-level monitoring for a job application.

The most effective integrity measure for any assessment is what happens after it: a technical follow-up conversation – conducted by SkillBrew.AI's AI Interview – that asks the candidate to explain their assessment responses, walk through their reasoning, or demonstrate the same skills in a different format.

A candidate who could not have produced their submitted answers independently will be unable to explain or extend them in a live follow-up. This post-assessment verification approach catches impersonation and AI-assistance with high accuracy while feeling like a natural part of the evaluation process rather than a punitive integrity check.

BrewShield is not a standalone proctoring product – it is an integrated integrity layer within SkillBrew.AI's full evaluation platform. Assessment integrity is meaningfully different from proctoring software sold as a standalone product: it is one dimension of a complete, orchestrated hiring evaluation system where the screening call, the assessment, the interview, and the post-assessment verification are all connected stages producing a coherent, consistent picture of the candidate.

See SkillBrew.AI's full assessment and integrity platform →