Exclusive: iMerit launches platform to connect AI developers with specialized domain experts
SiliconANGLE Paul Gillin
iMerit launched Scholars, a platform for finding experts to test and tune AI models. It targets the messy part of enterprise AI: catching mistakes humans still spot better than machines.
Based on reporting by SiliconANGLE, Paul Gillin — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
iMerit is trying to turn human expertise into a product. The data-annotation and AI training company, now part of EXL after its July acquisition, has launched iMerit Scholars, a platform for recruiting and managing specialists who can evaluate, fine-tune and validate AI models for enterprise use.
The pitch is aimed squarely at the awkward middle of AI development. Foundation models can get impressive fast, but iMerit says the harder job is adapting them to specific uses, especially in regulated fields like healthcare, banking and insurance. That’s where the company wants its experts to step in and, in Radha Basu’s phrase, “torment” the models until they fail in useful ways.
The network behind the platform is already large. iMerit says it has more than 30,000 vetted professionals across 118 countries, covering 106 languages and disciplines from medicine and mathematics to agriculture and autonomous systems. The company expects that pool to grow into the hundreds of thousands.
Scholars is built on iMerit’s Ango Hub platform and wraps recruiting, onboarding, evaluation, assignment and compensation into one system. It also supports multimodal training data, model evaluation and feedback workflows, while Trolley Technologies handles payment processing and compliance. The point is not just to label data, but to use specialists who can challenge models and catch errors automated checks might miss.
iMerit also says it is managing customer projects directly, rather than just matching clients with contractors. That means project management, quality assurance and auditing sit with the company, while experts remain part of a broader community and are paid per engagement. Pricing can be hourly, tied to deliverables or linked to outcomes, and Basu said the company can ramp quickly when demand spikes, including one case where a customer wanted about 400 extra workers in three or four days.
My take — AI-written commentary, not fact-checked reporting
This is the sane side of AI right now: less worship of giant models, more respect for people who can actually break them. The industry keeps pretending automation can audit itself, then acts shocked when that goes sideways. Human experts aren’t a nice-to-have here; they’re the bill that shows up after the demo.
Read more about this at: SiliconANGLE