EXL to acquire iMerit for $310m to deepen enterprise AI stack

EXL’s purchase of iMerit is a reminder that in the enterprise market, the moat isn’t just models—it’s data curation, evaluation pipelines, and human‑in‑the‑loop reinforcement learning. iMerit has long been one of the better‑known providers of annotation and RLHF‑style services for high‑stakes domains like healthcare, mobility, and robotics, backed by its Ango tooling platform and a distributed “Scholars” network of domain experts. ([exlservice.com](https://www.exlservice.com/about/newsroom/exl-to-acquire-imerit-advancing-its-leadership-in-enterprise-ai-by-adding-foundation-model-expertise-and-technology?utm_source=openai)) By bringing that stack in‑house and wiring it into EXL’s EXLerate.ai and related platforms, EXL is trying to offer clients a full pipeline: from raw proprietary data through labeling, model fine‑tuning, evaluation, and deployment.

For the AGI race, this is more about monetisation than frontier capability. But robust enterprise feedback loops matter: if companies like EXL can reliably turn messy domain data into tuned, evaluated models across regulated industries, that revenue funds further R&D and supplies real‑world safety and alignment data back to model builders. It also raises the competitive bar for other BPO and consulting players—simply reselling API access won’t cut it if rivals control the RLHF and evaluation layers. Over time, these enterprise‑grade alignment pipelines could influence how foundation labs think about safety‑critical deployments and long‑horizon agents in finance, health, and infrastructure.