> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mergeoa.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Merge: Assess Engineering Judgement

> Merge replaces algorithm quizzes with realistic pull request review loops, giving hiring teams evidence they can trust when evaluating engineers.

Merge is a code review assessment platform built for engineering hiring teams who want signal that reflects real work. Instead of whiteboard puzzles or disconnected coding exercises, Merge drops candidates into an authentic pull request loop — complete with a real codebase, scoped changes, and an AI that responds to their feedback in real time. The result is a structured, repeatable evaluation that tells you how a candidate actually thinks about code, not just whether they can reverse a linked list under pressure.

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="quickstart">
    Create your first assessment and send a candidate invite in minutes.
  </Card>

  <Card title="Assessments Overview" icon="clipboard-list" href="assessments/overview">
    Learn how to configure difficulty, specialization, and language surface.
  </Card>

  <Card title="How the Review Loop Works" icon="code-branch" href="review-loop/how-it-works">
    Understand the AI-driven pull request cycle candidates move through.
  </Card>

  <Card title="Integrations Overview" icon="plug" href="integrations/overview">
    Connect Merge to your ATS for seamless invite and result sync.
  </Card>
</CardGroup>

## How it works

Every Merge assessment follows the same four-step review loop. Candidates engage with real code, leave structured feedback, and watch the codebase evolve in response — giving your team a window into how they reason, prioritize, and communicate.

<Steps>
  <Step title="Read the codebase and pull request">
    Candidates receive access to a real codebase and a scoped pull request. They explore the diff, understand the context, and prepare to review the changes as they would on the job.
  </Step>

  <Step title="Comment on bugs, refactors, and vulnerabilities">
    Candidates leave inline comments identifying issues — whether that's a logic bug, a missed edge case, a refactoring opportunity, or a security risk. The depth and quality of these comments form the core of the evaluation.
  </Step>

  <Step title="AI publishes a revised pull request">
    Merge's AI reads the candidate's comments and publishes a fresh PR revision that addresses the feedback. The candidate can then review the new diff and respond again, just as they would in a real code review cycle.
  </Step>

  <Step title="Loop continues until time expires or the candidate approves">
    The review loop repeats — the candidate comments, the AI revises — until the assessment window closes or the candidate marks the PR approved. Every exchange is captured for scoring.
  </Step>
</Steps>

## Key features

Merge is designed to fit into how engineering teams already hire, with configuration options that match your roles and integrations that connect to the tools you already use.

<CardGroup cols={3}>
  <Card title="Custom Assessments" icon="sliders" href="assessments/overview">
    Set difficulty from intern to principal, choose a specialization, and pick the language surface that matches your open role.
  </Card>

  <Card title="ATS Integrations" icon="plug" href="integrations/overview">
    Sync candidate invites and completed scorecards directly with Lever, Greenhouse, Ashby, Workday, and more.
  </Card>

  <Card title="Scorecards" icon="chart-bar" href="reporting/scorecards">
    Every completed assessment produces a structured scorecard covering code quality reasoning, risk detection, revision judgment, and a hiring recommendation.
  </Card>

  <Card title="Proctoring" icon="shield" href="proctoring/overview">
    Optionally enable camera, microphone, and screen-share recording during the assessment window for added confidence in your results.
  </Card>

  <Card title="AI Revisions" icon="robot" href="review-loop/how-it-works">
    Merge's AI responds to candidate comments with a genuine PR revision, keeping the loop realistic and the evaluation consistent across every candidate.
  </Card>

  <Card title="Review Loop" icon="code-pull-request" href="review-loop/how-it-works">
    The multi-round review loop mirrors how engineers collaborate on real teams, surfacing communication quality alongside technical judgment.
  </Card>
</CardGroup>

<Note>
  Ready to see Merge in action? Book a 30-minute demo with the team [here](https://cal.com/mergeoademo/30min) to get access and run your first assessment. Thanks for your interest in Merge!
</Note>