acquisition-evaluator

Automated quality assurance for AI-generated cold emails applies weighted scoring across personalization, grammar, and CTA clarity. It acts as an independent acceptance layer, returning a Markdown-formatted audit report with actionable remediation steps to fix draft issues before release.

Agent output evaluation is an independent, standards-driven acceptance layer that scores AI-generated business outputs. By applying weighted checks and mandatory fail conditions, it catches low-quality or unsafe deliverables in validation workflows before they reach customers or downstream processes.

Enforcing mandatory quality gates for commercial quotations involves applying one-vote-fail rules to detect missing fields or cost leaks. The evaluator reviews generator outputs against strict pass/fail veto conditions, ensuring risky releases are blocked and flagged for remediation.

Automated quality control can audit due-diligence reports and WhatsApp messages through structured scoring across asset types. It evaluates content against specific weighted dimensions and produces an acceptance report detailing issues and remediation steps for manual QA or pipeline checkpoints.

The best way to integrate a post-stage evaluation checkpoint is using a coordinator-triggered validation workflow. A separate evaluator reviews generator outputs, applies dimension-specific scoring, and returns a Markdown audit report to trigger automated quality control before downstream processing.

A quotation generator fails the acceptance audit when outputs violate weighted dimension thresholds or trigger mandatory fail conditions. The evaluator applies strict one-vote-fail rules to detect subtle cost leaks or compliance issues, blocking the deliverable and requiring remediation before approval.