iMerit Launches Expert Network Platform for Enterprise AI Model Validation

ExlService Holdings Inc.'s iMerit subsidiary, acquired in July, has introduced a new platform aimed at linking artificial intelligence developers with domain-specific professionals capable of assessing, refining and certifying models destined for complex enterprise deployments. The iMerit Scholars platform manages a worldwide talent pool that the organization plans to expand to hundreds of thousands of specialists. Currently, the network comprises more than 30,000 qualified professionals spread across 118 countries, fluent in 106 languages and skilled in fields ranging from medicine and mathematics to agriculture and autonomous systems.

The initiative signals a fundamental shift in how artificial intelligence is being developed. Rather than focusing solely on creating more powerful foundational models, companies are now concentrating on customizing existing models for particular use cases, especially within heavily regulated sectors including healthcare, banking and insurance. According to Radha Basu, executive vice president and head of iMerit at EXL, "It's like having a product and saying, 'What are we going to do with it?'" She emphasized that specialized human knowledge is growing more critical as organizations implement AI in scenarios where incorrect outputs could trigger substantial harm.

The platform furnishes integrated capabilities for sourcing, bringing on staff, assessing, distributing work to and paying specialists. Built on iMerit's proprietary Ango Hub system for data annotation, process automation and model refinement, the solution facilitates teamwork on diverse data types, model assessment and iterative improvement processes. Trolley Technologies Inc. handles payment infrastructure and regulatory compliance through a partnership arrangement.

iMerit distinguishes itself from standard data labeling vendors by highlighting the capacity of its specialists to push back against AI systems and uncover flaws that computational methods might overlook. Basu described the methodology as follows: "We call it tormenting models. Question the model, question the model to make sure it's working, actually try to break it."

The organization employs a graduated screening framework to certify participants. An initial phase tests foundational AI knowledge and reasoning abilities. Subsequent phases assess expertise within specific domains and proficiency in collaborating with AI systems. The final stage involves assessments by fellow specialists within the community.

Rather than simply acting as a matchmaker between clients and freelancers, iMerit takes direct ownership of customer engagements. The company handles project oversight, quality monitoring and verification procedures. While specialists receive compensation per assignment, they maintain membership in the broader Scholars ecosystem. Basu noted that maintaining these ongoing connections proves valuable because experts accumulate understanding of particular models through repeated cycles of assessment and training.

iMerit provides multiple compensation structures to accommodate different project needs. Options include per-hour fees scaled to specialist qualifications, fixed-price agreements based on defined deliverables and performance-based arrangements tied to measurable results. Basu cited agricultural technology that forecasts crop losses early as an instance of work suitable for evaluation using performance metrics.

The platform also handles rapid changes in staffing requirements. Basu recounted a situation where a customer needed roughly 400 additional workers in a three to four day window to handle an unexpected surge in demand.

The ExlService ownership provides iMerit with deeper connections to healthcare and insurance sector knowledge, strengthening its position to serve industry-specific AI requirements. Basu indicated that expansion priorities center on deepening expertise across new specializations and languages. She stated: "Our biggest need right now is to go into new expertise, scale into newer languages and into newer levels of the same expertise."