EXL acquires AI data firm iMerit for $310 million
EXL has completed its acquisition of AI data specialist iMerit for up to $310 million, helping enterprise developers bridge the gap between experimental AI pilots and production.

ExlService Holdings, known as EXL, has finalized its acquisition of AI data specialist iMerit in a transaction valued at up to $310 million. The deal structure involves $170 million in upfront consideration alongside up to $140 million in performance-linked incentives and earnouts distributed over the next two years. Under the new arrangement, iMerit founder Radha Ramaswami Basu will transition to EXL as Executive Vice President and Head of iMerit, while also joining the firm's executive committee to help shape broader AI strategy alongside EXL Chairman and CEO Rohit Kapoor.
This acquisition directly merges iMerit's specialized model-development and evaluation tools with EXL's existing enterprise software suite. Specifically, iMerit's Ango Hub platform, which coordinates data annotation, reinforcement learning, and AI agent testing, will now sit alongside EXLerate.ai, an agentic AI platform utilizing NVIDIA technologies, and EXLdata.ai, a data-governance tool that integrates with AWS, Snowflake, and Databricks. By combining these technologies, the unified company can manage the entire lifecycle of an AI application, from initial unstructured data preparation to human-in-the-loop validation.
For enterprise AI practitioners, this consolidation addresses the difficult transition from a successful proof-of-concept to a reliable production system. Instead of managing separate vendors for data pipeline engineering, model fine-tuning, and safety testing, developers can leverage a single workflow. Through the iMerit Scholars network of domain experts, practitioners in highly regulated fields like healthcare, insurance, and banking gain access to specialized human feedback. This helps verify complex reasoning, conduct red teaming, and audit AI agent tool calls to prevent operational or regulatory failures.
This is our own summary of reporting by Unite.AI



