Knowing how to do a job can also qualify someone to judge whether AI has done it properly. But selling that judgment through a project platform is not the same as securing a permanent job.
Mercor’s November 10, 2026, announcement of a $750 million Series D at a $20 billion valuation brings that distinction into focus. The round included an employee tender, allowing employees to sell shares, so the headline amount should not be read as entirely new operating cash.
The Breakdown
Mercor connects subject specialists with work that helps train and test AI systems. The company says it is extending its evaluation and post-training capabilities from AI laboratories to enterprise customers.
Evaluation means defining a task and judging how well a system performs it. Post-training means further training an existing model for particular capabilities or behaviour. Neither is simply a matter of adding more documents.
Mercor’s approach includes tests based on a company’s actual workflows. This creates a role for people who understand both professional standards and the circumstances in which a particular answer would be useful.
That is the business proposition behind the funding, not proof that every profession will become an AI-training market.
An Indian example, with limits
A Mumbai-tagged Mercor recruitment listing for a remote India Domestic Tax Specialist illustrates the work. It asks experienced professionals to review Indian tax research and model answers, check statutory interpretation and supporting authorities, and refine grading criteria.
The listing specifies a remote, short-term contract and advertises $55–$80 an hour with a 10–20-hour weekly commitment. Those are advertised terms for that role, not verified earnings or a representative rate for Indian contractors. The listing is older than the funding announcement; it is an example of the work, not a newly announced hiring drive.
Mercor’s current payment documentation also includes India among supported countries. That establishes a payment route, not eligibility for every assignment or a guarantee of selection.
The economics are in the contract
Mercor’s documentation distinguishes hourly contracts from deliverable-based contracts. Hourly work is paid by tracked time and has a specified weekly cap. Deliverable-based work pays on successful completion of all specified items and has no minimum hourly guarantee.
For a contractor, the meaningful comparison is therefore not just the advertised rate. For hourly work, the rate, permitted hours and available assignments shape earnings. For deliverable-based work, completion requirements and the time needed to finish the agreed tasks determine the effective hourly return.
For buyers, the corresponding question is whether experts can turn professional judgment into clear, reusable assessment criteria. A tax answer that sounds convincing may still use the wrong authority or overlook a qualification. Checking those details is work in its own right.
The Crunch
Mercor’s funding is a bet on a market for human expertise inside AI development. Its Indian tax listing makes that opportunity concrete. Neither the valuation nor an advertised hourly rate establishes the scale, stability or typical earnings of that work in India.
Related categories: Technology & AI; Business & Brands
Leave your thoughts