Invisible Becoming Credentialed AI Gatekeeper

This points to Invisible becoming a gatekeeper for who is qualified to do high stakes AI work, not just a vendor that supplies labor. The WeCP deal adds testing infrastructure, a large bank of interview and assessment records, and role specific evaluation frameworks, which lets Invisible score and certify experts before assigning them to annotation, model eval, or agent supervision work inside Meridial. That shifts value from selling hours to selling trusted quality control.1 2 3 4

• Invisible already broke Meridial into modules for data prep, workflow building, expert sourcing, evaluation, and agent orchestration. Adding WeCP strengthens the verification layer, so expert selection becomes part of the product rather than a manual recruiting step bolted onto services.4 1
• The competitive pattern in AI data work is moving from bulk labeling toward smaller pools of vetted specialists. Prolific has expanded from participant matching into RLHF, annotation, and model evaluation with pre vetted raters, while Mercor is positioned around pre vetted domain experts for RLHF. Invisible can meet that shift by owning assessment and credentialing directly.5 6 7
• This also changes how revenue can be captured. Invisible has been charging for managed task execution through a distributed workforce, with estimated 2024 revenue of $134M. If certification, routing, and workflow controls become core software layers, more of each engagement can be priced like proprietary tooling instead of pure labor markup.3 8

The next step is a tighter loop where assessment, task routing, and performance history feed each other. That would let Invisible build a persistent credential graph for AI workers, use it to win more complex eval and safety programs, and push Meridial further toward a system of record for managed human and AI agent work.1 4 6