Invisible's Outcome-Based Managed Automation

Invisible wins by selling outcomes, not seats. A customer does not hire a bench of agents or buy a standalone software license. It hands over a messy workflow, like model evaluation, claims review, or onboarding operations, and Invisible breaks that work into steps, routes the repeatable parts through software, sends the judgment calls to people, and delivers the finished output with one owner and one bill.1 2 3 4

• This sits between classic outsourcing and pure SaaS. Traditional BPOs mostly add people to a process. Pure automation vendors sell tools and expect the customer to redesign work around them. Invisible wraps both together, which is why it could grow from about $2.5M in 2020 revenue to $134M in 2024 while staying profitable at roughly 11% EBITDA margin.1 2
• The mechanics are concrete. In AI training work, Invisible has used a 3,000 plus person global workforce and internal workflow software to split jobs into micro tasks, pay workers for the judgment layer, and charge customers more for the packaged service. That is closer to a managed production line than to a freelancer marketplace.3 4
• The closest comparables show why this middle position matters. Mercor is more of a talent marketplace, matching labs with experts. Prolific is more self serve, giving customers tools and APIs to run studies with its participant pool. Invisible is more hands on, taking responsibility for the end to end operation itself.4 5 6

The next step is deeper into regulated enterprise work. As synthetic data and better models reduce demand for generic labeling, the durable opportunity shifts to workflows where audit trails, human signoff, and process redesign still matter, especially in finance, insurance, healthcare, and other settings where software alone is not enough.3 4 6 7